Methods and compositions for protein expression and purification

Abstract

Methods for enhancing expression levels and secretion of heterologous fusion proteins in a host cell are disclosed.

Description

CROSS REFERENCE TO RELATED APPLICATION

This application is a continuation-in-part of U.S. application Ser. No. 10/338,411 filed Jan. 7, 2003, issued as U.S. Pat. No. 7,060,461, which claims priority to U.S. Provisional Application No. 60/346,449 entitled “Methods for Protein Expression and Purification” filed Jan. 7, 2002.

FIELD OF THE INVENTION

The present invention relates to the field of recombinant gene expression and purification of expressed proteins. More specifically, the invention provides materials and methods which facilitate purification of heterologous proteins from a variety of different host species.

BACKGROUND OF THE INVENTION

Several publications and patent documents are cited throughout the specification in order to describe the state of the art to which this invention pertains. Full citations for those references that are numbered can be found at the end of the specification. Each citation is incorporated herein as though set forth in full.

Functional genomic studies have been hampered by the inability to uniformly express and purify biologically active proteins in heterologous expression systems. Despite the use of identical transcriptional and translational signals in a given expression vector, expressed protein levels have been observed to vary dramatically (5, 7). For this reason, several strategies have been developed to express heterologous proteins in bacteria, yeast, mammalian and insect cells as gene-fusions.

The expression of heterologous genes in bacteria is by far the simplest and most inexpensive means available for research or commercial purposes. However, some heterologous gene products fail to attain their correct three-dimensional conformation in E. coli while others become sequestered in large insoluble aggregates or “inclusion bodies” when overproduced. Major denaturant-induced solubilization methods followed by removal of the denaturant under conditions that favor refolding are often required to produce a reasonable yield of the recombinant protein. Selection of ORFs for structural genomics projects has also shown that only about 20% of the genes expressed in E. coli render proteins that were soluble or correctly folded (36, 38). These numbers are startlingly disappointing especially given that most scientists rely on E. coli for initial attempts to express gene products. Several gene fusion systems such as NUS A, maltose binding protein (MBP), glutathione S transferase (GST), and thioredoxin (TRX) have been developed (17). All of these systems have certain drawbacks, ranging from inefficient expression to inconsistent cleavage from desired structure. Comprehensive data showing that a particular fusion is best for a certain family of proteins is not available.

Ubiquitin and ubiquitin like proteins (UBLs) have been described in the literature. The SUMO system has also been characterized. SUMO (small ubiquitin related modifier) is also known as Sentrin, SMT3, PIC1, GMP1 and UBL1. SUMO and the SUMO pathway are present throughout the eukaryotic kingdom and the proteins are highly conserved from yeast to humans (12, 15, 28). SUMO homologues have also been identified in C. elegans and plants. SUMO has 18% sequence identity with ubiquitin (28, 39). Yeast has only a single SUMO gene, which has also been termed SMT3 (23, 16). The yeast Smt3 gene is essential for viability (29). In contrast to yeast, three members of SUMO have been described in vertebrates: SUMO-1 and close homologues SUMO-2 and SUMO-3. Human SUMO-1, a 101 amino-acid polypeptide, shares 50% sequence identity with human SUMO-1/SUMO-2 (29). Yeast SUMO (SMT3) shares 47% sequence identity with mammalian SUMO-1. Although overall sequence homology between ubiquitin and SUMO is only 18%, structure determination by nuclear magnetic resonance (NMR) reveals that the two proteins share a common three dimensional structure that is characterized by a tightly packed globular fold with β-sheets wrapped around one α-helix (4). Examination of the chaperoning properties of SUMO reveals that attachment of a tightly packed globular structure to N-termini of proteins can act as nucleus for folding and protect the labile protein. All SUMO genes encode precursor proteins with a short C-terminal sequence that extends from the conserved C-terminal Gly-Gly motif. The extension sequence, 2–12 amino acids in length, is different in all cases. Cells contain potent SUMO proteases that remove the C-terminal extensions. The C-terminus of SUMO is conjugated to E amino groups of lysine residues of target proteins. The similarity of the enzymes of the sumoylation pathway to ubiquitin pathway enzymes is remarkable, given the different effects of these two protein modification pathways. Sumoylation of cellular proteins has been proposed to regulate nuclear transport, signal transduction, stress response, and cell cycle progression (29). It is very likely that SUMO chaperones translocation of proteins among various cell compartments, however, the precise mechanistic details of this function of SUMO are not known.

Other fusions promote solubility of partner proteins presumably due to their large size (e.g., NUS A). Fusion of proteins with glutathione S-transferase (GST) or maltose binding protein (MBP) has been proposed to enhance expression and yield of fusion partners. However, enhanced expression is not always observed when GST is used as GST forms dimers and can retard protein solubility. Another problem with GST or other fusion systems is that the desired protein may have to be removed from the fusion. To circumvent this problem, protease sites, such as factor X, thrombin or Tev protease sites are often engineered downstream of the fusion partner. However, incomplete cleavage and inappropriate cleavage within the fusion protein is often observed. The present invention circumvents these problems.

SUMMARY OF THE INVENTION

In accordance with the present invention compositions and methods for enhancing expression levels of a protein of interest in a host cell are provided. An exemplary method comprises i) operably linking a nucleic acid sequence encoding molecule selected from the group consisting of SUMO, RUB, HUB, APG8, APG12, URM1, and ISG15 to a nucleic acid sequence encoding said protein of interest thereby generating a construct encoding a fusion protein, ii) introducing said nucleic acid into said host cell, whereby the presence of said molecule in said fusion protein increases the expression level of said protein of interest in said host cell. In a preferred embodiment the molecule is SUMO encoded by a nucleic acid of SEQ ID NO: 2. The method optionally entails cleavage of said fusion protein and isolation of the protein of interest.

In yet another embodiment of the invention, an exemplary method for generating a protein of interest having an altered amino terminus is provided. Such a method comprises i) providing a nucleic acid sequence encoding the protein of interest; ii) altering the N-terminal amino acid coding sequence in the nucleic acid; iii) operably linking a SUMO molecule to the nucleic acid sequence; and iv) expressing the nucleic acid in a eukaryotic cell, thereby producing the protein of interest in the cell, wherein the eukaryotic cell expresses endogenous SUMO cleaving enzymes, which effect cleavage of SUMO from the sequence encoding the protein of interest, thereby producing a protein of interest having an altered amino terminus. All amino acids with the exception of proline may be added to the amino terminus using this method.

The invention also provides a method for producing a sumolated protein for tracking protein localization within a host cell. An exemplary method comprises i) providing a nucleic acid sequence encoding said protein; ii) substituting the N-terminal amino acid coding sequence in the nucleic acid for a codon which encodes proline; iii) operably linking a SUMO molecule to said nucleic acid sequence; and iv) expressing said SUMO linked protein in said host cell.

In another aspect of the invention, a method for enhancing secretion levels of a protein of interest from a host cell is provided. Such a method comprises i) operably linking a nucleic acid sequence encoding molecule selected from the group consisting of SUMO, RUB, HUB, URM1, and ISG15 to a nucleic acid sequence encoding said protein of interest thereby generating a construct encoding a fusion protein, ii) introducing said nucleic acid into said host cell, whereby the presence of said molecule in said fusion protein increases the secretion of said protein of interest from said host cell.

In yet a further aspect of the invention, kits are provided for performing the methods described above. Such kits comprise a recombinant vector containing a nucleic acid sequence encoding a UBL molecule selected from the group of SUMO, RUB, HUB, URMl, and ISGl5 operably linked to a promoter suitable for expression in the desired host cell and a multiple cloning site suitable for cloning a nucleic acid encoding the protein of interest. The recombinant vector may also contain a nucleic acid sequence encoding for a purification tag. The kits may further comprise a preparation of a protease capable of cleaving the UBL molecule from the fusion protein, an appropriate solid phase for binding the purification tag, appropriate buffers including wash and cleavage buffers, and frozen stocks of host cells. The host cells may be selected from the group of yeast cells, E. coli, insect cells, and mammalian cells.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a schematic drawing illustrating the conjugation pathways for ubiquitin and ubiquitin-like proteins (UBLs). An arrow in the “C-terminal hydrolase” column indicates the cleavage of the precursor proteins. Only enzymes previously described are provided. The failure to list a particular enzyme in a particular pathway does not preclude the existence of that enzyme.

FIG. 2 is a schematic representation of the cloning strategy used to express SUMO tusion proteins. In this cloning strategy, a Bsa I site is introduced directly downstream of a SUMO sequence within a desired vector. The nucleic acid sequence encoding the protein to be expressed as a fusion with SUMO is amplified by PCR with primers that introduce a Bsa I site at the 5′ end. The vector (SEQ ID NO: 62, top strand; SEQ ID NO: 63, bottom strand) and the PCR product (SEQ ID NO: 60, top strand; SEQ ID NO: 61, bottom strand) are cleaved by Bsa I and an appropriate restriction enzyme (represented by Xxx) that allows for insertion of the cleaved PCR product into the vector.

FIG. 3 is a circular map of pSUMO, an E. coli SUMO expression vector. The nucleic acid sequence provided (SEQ ID NO: 37) encompasses the SUMO encoding region and the multiple cloning site. The amino acid sequence provided (SEQ ID NO: 38) is 6×His tagged SUMO. Restriction enzymes are indicated above their recognition sequence. The pSUMO expression vector has been constructed in the backbone of the pET-24d expression vector (Novagen).

FIGS. 4A and 4B show Coomassie stained gels and graphic data that demonstrate that the attachment of the carboxy-terminus of UBLs to the amino-terminus of target proteins increases expression and/or enhances solubility of the protein in E. coli. Green fluorescence protein (GFP) and UBL-GFP fusions encoded in pET24d E. coli expression vectors were expressed in the E. coli Rosetta pLysS strain (Novagen). Expression was induced either at 37° C. with 1 mM IPTG for four hours either in LB medium (FIG. 4A) or in minimal media with 1 mM IPTG at 26° C. overnight (FIG. 4B). Left panels are Coomassie stained SDS-polyacrylamide gels of total cellular protein (top) and soluble proteins (bottom). The first lanes of each gel are molecular weight markers. Dark arrow indicates observed GFP species and light arrow indicates size of expected GFP species. Right panel is quantitative representation in Arbitrary Units (AU) of GFP fluorescence present in soluble fractions as measured in a Fluorscan Ascent FL fluorometer (LabSystems).

FIG. 5 is a Coomassie stained SDS-polyacrylamide gel demonstrating the expression and purification of a human tyrosine kinase as a SUMO fusion protein in E. coli. Tyrosine kinase and the fusion protein SUMO-tyrosine kinase were expressed in the Rossetta pLysS strain (Novagen) of E. coli in LB or minimal media (MM). The right panel shows the Ni-NTA resin purified proteins from the transformed E. coli cells. The left panel has the same lane arrangement as the right panel, but ⅓ of the amount protein was loaded on the SDS-polyacrylamide gel. Numbers indicate molecular weight standards in the first lane.

FIG. 6 shows a Coomassie stained SDS-polyacrylamide gel representing purified SUMO hydrolase from E. coli and the partial purification and elution of SUMO-tyrosine kinase fusion protein. E. coli cells were transformed with a vector expressing either SUMO hydrolase Ulp1 or SUMO-tyrosine kinase and cultured in minimal media. Proteins were subsequently purified by Ni-NTA resin. SUMO-tyrosine kinase was further purified by elution with either 100 mM EDTA or 250 mM imidazole. The gel shows that the current methods yield approximately 90% pure Ulp1 protein.

FIG. 7 is a stained SDS-polyacrylamide gel of the expression of the liver X receptor (LXR) ligand binding domain as a fusion protein with SUMO. E. coli cells were transformed with a SUMO-LXR expression vector. The cells were subsequently induced with 1 mM IPTG at 20° C. overnight or 37° C. for 3 hours. 10 μg of total protein (WC), soluble protein (CS), and insoluble protein (Insol) from each induction were loaded per well of a 12% SDS-polyacrylamide gel.

FIGS. 8A and 8B display stained SDS-polyacrylamide gels demonstrating the solubility of the SUMO-MAPKAPK2 fusion protein expressed at 37° C. (FIG. 8A) and 20° C. (FIG. 8B). E. coli cells expressing a SUMO-fusion of MAPKAP2 kinase were induced with 0.1 (lanes 2–4), 0.25 (lanes 5–7), and 0.5 (lanes 8–10) mM IPTG. The original induction sample (I) in addition to the supernatant (S) and resuspended pellet (P) following lysis and centrifugation were analyzed by SDS-PAGE. The first lanes are BioRad low molecular weight markers.

FIG. 9 is a Western blot (top panel) of UBL-GFP fusion proteins expressed in yeast cells demonstrating that UBL-GFP fusion proteins are co-translationally cleaved in yeast. Yeast strain BJ1991 was transformed with a vector expressing Ub-GFP, SUMO-GFP, Urm1-GFP, Hub1-GFP, Rub1-GFP, Apg8-GFP, Apg12-GFP or ISG15-GFP under the control of a copper sulfate regulated promoter. Total cell extracts were prepared by boiling the cells in SDS-PAGE buffer and briefly sonicating the sample to reduce viscosity. 20 μg of the total yeast proteins were resolved on 12% SDS-PAGE minigels and analyzed by Western blot with a rabbit polyclonal antibody against GFP and a secondary HRP-conjugated antibody. The arrow indicates the size of unfused GFP. An identical gel (bottom panel) was run in parallel and stained with Coomassie to ensure equal loading of the proteins from all samples.

FIG. 10 is a series of Western blots that indicate SUMO-GFP Fusions are co-translationally cleaved in yeast generating novel amino termini. In addition to methionine as the first amino acid of GFP following the C-terminal Gly-Gly sequence of SUMO, we have engineered the remaining 19 amino acids as the amino-terminal residue of GFP in yeast SUMO-(X)20-GFP expression vectors. All expression vectors containing the 20 amino-terminal variants of GFP fusion proteins were expressed in yeast under the control of copper inducible promoter. Yeast lysates were separated by SDS-PAGE and analyzed by Western blot with antibodies against GFP. The “unfused-GFP” lanes represent the expression of GFP alone with no SUMO fusion. The “SUMO-GFP” lanes are bacterially expressed SUMO-GFP.

FIGS. 11A and 11B are schematic representations of the SUMO (FIG. 11A) and ubiquitin (FIG. 11B) GFP fusion proteins that also contain the gp67 secretory signal. In construct E, only unfused GFP protein is expressed. In construct G, a 7 kDa secretory sequence from gp67 was attached to the N-terminus of GFP. In constructs S and U, SUMO and ubiquitin sequences, respectively, are inserted in frame to the N-terminus of GFP. In constructs GS and GU, gp67 sequences are followed by SUMO and ubiquitin, respectively, and then GFP. In constructs SG and UG, gp67 sequences are inserted in between the C-terminus of SUMO and ubiquitin, repectively, and the N-terminus of GFP.

FIGS. 12A and 12B are Western blots demonstrating expression of SUMO and ubiquitin fusion proteins in insect cells. Hi-five insect cells were infected with recombinant baculovirus encoding for SUMO or ubiquitin fusion proteins. At 24 hours post-infection, equal amounts of cell lysates (FIG. 12A) and media (FIG. 12B) were separated by SDS-PAGE and analyzed by Western blot with antibodies against GFP. Lane markers: Hi5 is Hi Five cells, E is eGFP, G is gp67-eGFP, U is ubiquitin-eGFP, S is SUMO-eGFP, GU is gp67-ubiquitin-eGFP, UG is ubiquitin-gp67-eGFP, GS is gp67-SUMO-eGFP, SG is SUMO-gp67-eGFP, and eGFP is a positive control.

FIGS. 13A, 13B, and 13C are Western blots demonstrating expression of SUMO and ubiquitin fusion proteins in insect cells. Hi-five insect cells were infected with recombinant baculovirus encoding for SUMO or ubiquitin fusion proteins. At 48 hours post-infection, equal amounts of cell lysates (FIGS. 13A and 13C) and media (FIG. 13B) were separated by SDS-PAGE and analyzed by Western blot with antibodies against GFP. The lanes are: Hi5 is Hi Five cells, E is eGFP, G is gp67-eGFP, U is ubiquitin-eGFP, S is SUMO-eGFP, GU is gp67-ubiquitin-eGFP, UG is ubiquitin-gp67-eGFP, GS is gp67-SUMO-eGFP, SG is SUMO-gp67-eGFP, and S-P is SUMO-proline-GFP.

FIG. 14 is a series of micrographs of eGFP expression in Hi-Five cells infected with different eGFP fusion baculoviruses. Pictures were taken with a Leitz Fluovert Inverted Microscope with excitation at 488 nm with Hammamatsu Orca Cooled CCD camera.

FIG. 15 contains stained SDS-polyacrylamide gels representing the in vitro Ulp1 cleavage of Ni-NTA resin purified His6SUMO-eGFP fusion proteins expressed in E. coli. The purified His6SUMO-eGFP fusions, containing a different amino acid at the +1 position of the Ulp1 cleavage site, were incubated at 30° C. for 3 hours with purified Ulp1 hydrolase. The lanes are marked with the single letter code of the +1 amino acid. The negative control (−Ve) is the incubation of His6SUMO-eGFP at 30° C. for 3 hours in the absence of enzyme. Low molecular weight markers (LMW) are also provided.

FIG. 16 contains a pair of stained SDS-polyacrylamide gels representing the effects of various conditions on Ulp1. Ni-NTA purified His6SUMO-GFP was incubated with Ulp1 under the indicated conditions for one hour at room temperature unless indicated otherwise. Low molecular weight markers (LMW) are also provided.

FIG. 17 is a stained SDS-polyacrylamide gel representing the effects of various protease inhibitors on Ulp1. Ni-NTA purified His6SUMO-GFP was incubated with Ulp1 and 10 mM of various protease inhibitors for 1 hour at room temperature. Lane markers: Norm is addition of Ulp1 and N-ethymaleimide (NEM) to the substrate at the same time, Pre is the incubation of Ulp1 with NEM prior to the addition of substrate, +Ve is the absence of any inhibitor, −Ve is in the absence of Ulp1, lane 1 is with E-64, lane 2 is with EDTA, lane 3 is with leupeptin, lane 4 is with NEM, lane 5 is with pepstatin, lane 6 is with TLCK. Low molecular weight markers (LMW) are also provided.

FIG. 18 is a stained SDS-polyacrylamide gel showing purification and cleavage of MAPKAP2. E. coli transformed with the expression vector for SUMO-MAPKAP2 where either grown at 37° C. and induced with 0.1 mM IPTG (lanes 2–7) or at 20° C. and induced with 0.5 mM IPTG (lanes 8–13). Cell lysates were Ni-NTA purified and separated by SDS-PAGE. Lane 1: BioRad low molecular weight marker; lanes 2 and 8: soluble fraction of cell lysates; lanes 3 and 9: flow through from Ni-NTA column; lanes 4 and 10: 15 mM imidazole wash of Ni-NTA column; lanes 5 and 11: 300 mm imidazole elution of Ni-NTA column; lanes 6 and 12: supernatant of 2 hour incubation of elution with SUMO hydrolase at 30° C.; and lanes 7 and 13: pellet of hydrolase incubation.

FIG. 19 is a stained SDS-polyacrylamide gel showing SUMO hydrolase function at pH 7.5 and 8.0. Purified SUMO-GFP was cleaved using 1/50 diluted purified stock of SUMO hydrolase in sodium phosphate buffer pH 7.5 (lanes 1–6) and 8.0 (lanes 8–13) at room temperature for the following length of times: lanes 1 and 8: 0 minutes, lanes 2 and 9: 1 min, lanes 3 and 10: 2.5 min, lanes 4 and 11: 5 min, lanes 5 and 12: 10 min, and lanes 6 and 13: 20 min. Lane 7 is blank and M is molecular weight markers.

FIG. 20 is a stained SDS-polyacrylamide gel indicating SUMO hydrolase cleaves SUMO-β-Galactosidase. Purified SUMO hydrolase was incubated with E. coli produced SUMO-β-Galactosidase at room temperature for 0 minutes (lane 1), 2.5 min (lane 2), 5 min (lane 3), 10 min (lane 4), and 20 min (lane 5). Molecular weight markers are provided in lane M.

FIG. 21 is a stained SDS-polyacrylamide gel showing the cleavage of SUMO-GUS by SUMO Hydrolase in the presence of urea. Ni-NTA purified SUMO-β-GUS was incubated with 1/50 dilution of purified stock of SUMO hydrolase for 1 hour in increasing concentrations of urea at pH 8.0. Lane markers: M is broad range molecular weight marker; lane 1 is SUMO-GUS from soluble E. coli fraction; lane 2: flow through from nickel column; lane 3: wash; lane 4: elution; lanes 5–9: SUMO-GUS and hydrolase with various denaturants, specifically, lane 5: none; lane 6: 1 mM DTT; lane 7: 0.5 M Urea; lane 8: 1.0M Urea; lane 9: 2.0M Urea.

FIG. 22 is a stained SDS-polyacrylamide gel demonstrating the rapid isolation of a SUMO fusion protein. E. coli cells expressing a single IgG binding domain from Protein G fused to His6Smt3 were lysed with guanidinium chloride lysis buffer. Cell lysate supernatants were purified over Ni-NTA and eluted in a native buffer that allows for cleavage by Ulp1. Lane markers: PMW is molecular weight markers; lane 1 is cellular proteins prior to treatment with guanidinium chloride, lane 2 is guanidinium chloride cell lysates, lane 3 is flow through from Ni-NTA column, lane 4 is elution, and lane 5 is Ulp1 cleavage of elution.

FIG. 23 is the amino acid (SEQ ID NO: 1) and nucleotide (SEQ ID NO: 2) sequences of SUMO.

FIGS. 24A and 24B are the amino acid (SEQ ID NO: 3) and nucleotide (SEQ ID NO: 4) sequences of GFP.

FIGS. 25A and 25B are the amino acid (SEQ ID NO: 5) and nucleotide (SEQ ID NO: 6) sequences of SUMO-GFP.

FIGS. 26A and 26B are the amino acid (SEQ ID NO: 7) and nucleotide (SEQ ID NO: 8) sequences of ubiquitin-GFP.

FIGS. 27A and 27B are the amino acid (SEQ ID NO: 9) and nucleotide (SEQ ID NO: 10) sequences of URM1-GFP.

FIGS. 28A and 28B are the amino acid (SEQ ID NO: 1) and nucleotide (SEQ ID NO: 12) sequences of HUB1-GFP.

FIGS. 29A and 29B are the amino acid (SEQ ID NO: 13) and nucleotide (SEQ ID NO: 14) sequences of RUB1-GFP.

FIGS. 30A and 30B are the amino acid (SEQ ID NO: 15) and nucleotide (SEQ TD NO: 16) sequences of APG8-GFP.

FIGS. 31A and 31B are the amino acid (SEQ ID NO: 17) and nucleotide (SEQ ID NO: 18) sequences of APG12-GFP.

FIGS. 32A and 32B are the amino acid (SEQ ID NO: 19) and nucleotide (SEQ ID NO: 20) sequences of ISG15-GFP.

FIG. 33 is the amino acid (SEQ ID NO: 21) and nucleotide (SEQ ID NO: 22) sequences of SUMO-Protein G.

FIGS. 34A, 34B, and 34C are the amino acid (SEQ ID NO: 23) and nucleotide (SEQ ID NO: 24) sequences of SUMO-β GUS.

FIGS. 35A, 35B, and 35C are the amino acid (SEQ ID NO: 25) and nucleotide (SEQ ID NO: 26) sequences of SUMO-LXRα.

FIGS. 36A and 36B are the amino acid (SEQ ID NO: 27) and nucleotide (SEQ ID NO: 28) sequences of SUMO-Tyrosine Kinase.

FIGS. 37A and 37B are the amino acid (SEQ ID NO: 29) and nucleotide (SEQ ID NO: 30) sequences of SUMO-MPAKAP2 Kinase.

FIGS. 38A, 38B, 38C, 38D, and 38E are the amino acid (SEQ ID NO: 31) and nucleotide (SEQ ID NO: 32) sequences of SUMO-β GAL.

FIG. 39 is a circular map of YEpSUMO-eGFP.

FIGS. 40A, 40B, 40C, 40D, 40E and 40F are the nucleotide sequence (SEQ ID NO: 33) of YEpSUMO-eGFP. Select restriction enzyme sites are indicated.

FIG. 41 is a circular map of YEpUbGUS.

FIGS. 42A, 42B, 42C, 42D, 42E, 42F, and 42G are the nucleotide sequence (SEQ ID NO: 34) of YEpUbGUS. Select restriction enzyme sites are indicated.

FIG. 43 is a circular map of pFastBac SUMO-eGFP.

FIGS. 44A, 44B, 44C, 44D, and 44E are the nucleotide sequence (SEQ ID NO: 35) of pFastBac SUMO-eGFP. Select restriction enzyme sites are indicated.

FIG. 45 is a circular map of pSUMO (pET24d6His×SUMO).

FIGS. 46A, 46B, 46C, 46D, and 46E are the nucleotide sequence (SEQ ID NO: 36) of pSUMO (pET24d6His×SUMO). Select restriction enzyme sites are indicated.

DETAILED DESCRIPTION OF THE INVENTION

There are a number of reasons for the lack of efficient recombinant protein expression in a host, including, for example, short half life, improper folding or compartmentalization and codon bias. While the Human Genome project has successfully created a DNA “map” of the human genome, the development of protein expression technologies that function uniformly in different expression platforms and for all the protein motifs has not yet been achieved.

In accordance with the present invention, it has been discovered that that N-terminal fusion of the ubiquitin homologue SUMO or Smt3 to otherwise unexpressed or poorly expressed proteins remarkably enhances the expression levels of biologically active proteins in both prokaryotes and eukaryotes. The Ubiquitin-Like protein (UBL) family contains many proteins, including for example, SUMO, Rub1, Hub1, ISG15, Apg12, Apg8, Urm1, Ana 1a and Ana 1b (15, 28). See Table 1. The hallmark of all of these proteins, exept APG12, and URM1, is that they are synthesized as precursors and processed by a hydrolase (or proteases) to generate mature carboxy-terminal sequence. Secondly, all of the UBLs share a common structure.

In E. coli, fusion proteins remained intact while in yeast or insect cells fusion proteins were efficiently cleaved, except when proline was the N-terminal residue of the target protein. While any of the UBLs set forth in Table 1 may be utilized in the compositions and methods of the invention to enhance expression of heterologous fusion proteins of interest, SUMO is exemplified in the gene fusion system provided herein.

TABLE 1
Properties of Ubiquitin-like Proteins (UBLs)
UBL		Knockout		% UB		Hydro-	COOH
(yeast)	Function	phenotype	Substrate	Identity	KDa	lase	Residues
UB	Translocation	not viable	many	100	8.5	UGH/U	LRLR
	to					BPs	GG
	proteasome						(SEQ ID
	for						NO: 39)
	degradation.
SUMO	Translocation	not viable	Sentrins,	18	11.6	Aut1/Aut2	GG
(SMT3)	to nucleus		RanGap,
			others
RUB1	Regulation of	viable;	cullins,	60	8.7	not	GG
(NEDD8)	mitosis.	non-	cytoskelet			known
		essential.	proteins
HUB1	Cell	viable;	Sph1,	22	8.2	not	YY
	polarization	deficient in	Hbt1 cell			known
	during	mating.	polarity
	mating		factors
	projections.
ISG-15	Unknown	IFN, LPS	many	~30; 28	15.0	UBP43	LRLR
(UCRP)		hypersensi-		(two		(USP18)	GG (SEQ
		tivity; death		domains)			ID NO:
							39)
APG12	Autophagy	viable,	Apg5	18	21.1	not	FG
		defective in				cleaved
		autophagy
URM1	Unknown	ts growth;	unknown	20	11.0	not	GG
		non-				known
		essential.
APG8	Autophagy	viable; no	phospatid	18	13.6	Apg4/Aut2	FG
(LC3)		autophago-	yl-
		cytosis or	ethanol-
		sporulation	amine

The SUMO fusion system of the present invention has been successfully applied to express different molecular weight proteins such as 6 KDa Protein G domain to 110 KDa β-galactosidase in E. coli and eukaryotic cells. More specifically, the system allows one to: (1) enhance the expression of under-expressed proteins; (2) increase the solubility of proteins that are insoluble; (3) protect candidate proteins from degradation by intracellular proteases by fusing UBLs to their N-termini; (4) cleave the fusion protein to efficiently generate authentic proteins using naturally-present enzymes (5) generate proteins with novel amino termini; and (6) cleave all fusion proteins with remarkable efficiency irrespective of the N-terminal sequence of the fused protein, using UBL hydrolases such as SUMO hydrolase Ulp1. Because UBLs are small molecular weight proteins (˜100 amino acids), they can also be used as purification tags as well. These remarkable properties of UBLs make them excellent candidates for enhancing expression and solubility of proteins. The method may also be utilized to generate novel amino termini on proteins of interest for a variety of research, diagnostic and therapeutic applications.

The ultimate fate of ubiquitinated or sumoylated proteins within a cell varies. A protein can be monoubiquitinated or polyubiquitinated. Ubiquitination of protein has multiple functions and gives rise to different fates for the protein within a cell (11). Ubiquitination primarily targets proteins to 26S proteosome for degradation (13). On the other hand, sumoylation of target proteins does not lead to degradation, but, rather, leads directly or indirectly to altered localization of proteins (15). There are about 17 deubiquitinating enzymes that cleave conjugated ubiquitin from target proteins as well as ubiquitin-ubiquitin and ubiquitin artificial-fusion proteins (1, 35). Thus far it appears that yeast has two cysteinyl proteases, called Ulp1 and Ulp2, that remove SUMO from ε-amino groups of lysine as well from the artificial linear SUMO-fusions (20, 21).

To determine if UBLs and SUMO fusion will enhance expression of recombinant proteins of different sizes and function, we have designed several UBL-GFP fusion proteins in addition to SUMO-fusion proteins and monitored their expression levels in E. coli, yeast and insect cells. In E. coli, the proteins are expressed as intact fusions, while in eukaryotes, the fusions were efficiently cleaved. A dramatic increase in the yield of proteins after fusion with SUMO and expression in E. coli was observed. In additional studies, SUMO-GFP protein was used as a model fusion for detailed studies in yeast and insect cells. We have designed SUMO-GFP fusion where all the N-terminal methionine residues have been replaced with the rest of the 19 amino acids. We have purified 20 sumo-GFP fusion proteins from E. coli and cleaved them in vitro with Ulp1. Ulp1 efficiently cleaved 19 out of the 20 possible amino acid junctions. The proline junction was not cleaved. As compared to deubiquitinating enzyme (3), Ulp1 demonstrated broad specificity and robustness in its digestion properties. Proteins having a wide range of molecular weights were cleaved efficiently by Ulp1. Similarly, in yeast, and insect cells, the fusion proteins were efficiently processed, yielding intact, biologically active proteins. In addition to enhancing protein expression levels, the SUMO-fusion approach can be used to advantage to generate desired N-termini to study novel N-terminal protein functions in the cell. Since SUMO fusion can both enhance recombinant protein yield and generate new N-termini, this technology provides an important tool for post-genomic biotechnology analyses.

The present invention also encompasses kits for use in effecting enhanced expression, secretion, purification, localization, and alteration of the amino terminus of a protein of interest. Such kits comprise a recombinant vector containing a nucleic acid sequence encoding a UBL molecule selected from the group of SUMO, RUB, HUB, URM1, and ISG15 operably linked to a promoter suitable for expression in the desired host cell and a multiple cloning site suitable for cloning a nucleic acid encoding the protein of interest in-frame with the nucleic acid sequence encoding the UBL molecule. The promoter is preferably a strong promoter and may be constitutive or regulated. Such promoters are well known in the art and include, but are not limited to, the promoters provided hereinbelow such as the ADH1, T7, and CUP1 promoters.

The recombinant vector may also contain a nucleic acid sequence encoding a purification tag in-frame with the sequence encoding the UBL molecule. Purification tags are well known in the art (see Sambrook et al., 2001, Molecular Cloning, Cold Spring Harbor Laboratory) and include, but are not limited to: polyhistidine, glutathione-S-transferase, maltose binding protein, thioredoxin, the FLAG™ epitope, and the c-myc epitope. Materials and methods for the purification of fusion proteins via purification tags are also well known in the art (see Sambrook et al., Novagen catalog, 2002, examples hereinbelow). Reagents including, but not limited to, solid supports capable of binding the purification tag, lysis buffers, wash buffers, and elution buffers may also be included in the kits.

The kits may further comprise a composition comprising a protease or proteases capable of cleaving the UBL molecule from the fusion protein, cleavage buffers, frozen stocks of host cells, and instruction manuals. The kits may also further comprise reagents for altering the nucleic acid encoding a protein of interest to generate amino termini which are different from those native to the wild-type protein. Methods for altering the nucleic acid are well known in the art and include, but are not limited to, site-directed mutagenesis and oligonucleotide-based site-directed mutagenesis (see BD Biosciences Catalog, 2001; Qiagen Catalog, 2001; Ausubel et al., eds., 1995, Current Protocols in Molecular Biology, John Wiley and Sons, Inc.).

As used herein, an “instructional material” includes a publication, a recording, a diagram, or any other medium of expression which can be used to communicate the usefulness of the composition of the invention for performing a method of the invention. The instructional material of the kit of the invention can, for example, be affixed to a container which contains a kit of the invention to be shipped together with a container which contains the kit. Alternatively, the instructional material can be shipped separately from the container with the intention that the instructional material and kit be used cooperatively by the recipient.

The materials and methods set forth below are provided to facilitate the practice of the present invention.

Design and Construction of E. coli Expression Vectors:

The original vector backbone was developed using pET 24d vector from Novagen (see FIG. 3 as well as FIGS. 45–46A–E). pET24d uses a T7 promoter system that is inducible with IPTG. The vector has a kanamycin selection marker and does not contain any translation terminator.

Construction of Variable His6SUMO-GFP Fusions:

A N-terminal six his-tagged SUMO (fusion vector was constructed as follows. A PCR product was generated with the primers 5 ′ CCATGGGTCATCACCATCATCATCACGGGTCGGACTCAGAAGTCAATCAA- 3 ′ (SEQ ID NO: 40) and 5′-GGATCCGGTCTCAACCTCCAATC TGTTCGCGGTGAG-3′(SEQ ID NO:41) using yeast Smt3 gene (16) as a template (kind gift of Erica Johnson). The FCR fragment was double digested with Nco I and Bam HI, and then ligated into pET24d, which had been similarly digested. It is important to note that the current invention utilizes a variant of the wild type yeast SUIvIO sequence. The A nucleotide at position 255 has been replaced with a G nucleotide, thus encoding an alanine instead of a threonine (SEQ ID NOS: 64 and 65). The detailed cloning strategy is provided in FIG. 2. The pET24d His6Smt3eOFP fusions, containing each of the twenty different amino acids at the +1 position of the cleavage site were generated as follows. The eGFP sequence was amplified a template, with the primers 5′-GGTCTCAAGGT NNNGTGAGCAAGGGCGAGGAGC-3′(SEQ ID NO:42) and 5′-AAGCTTATTACTTGTACAGCTCGT CCATGCC-3(SEQ ID NO: 43), where the NNN in the forward primer corresponding to the variable codon encoding one of the twenty amino acids. The PCR products were purified and double digested with Bsa I and Hind III, these were then ligated into the pET24dI-IisSUMO vector which had been similarly digested. Plasmids from clones containing the variable inserts, were sequenced to confirm the presence of the novel codon in each.

Construction of SUMO-Fusion Vectors from pSUMO:

The gene encoding the protein of interest is cloned in frame with the SUMO tag, in the pSUMO vector, by utilizing the encoded Bsa I site. Bsa I belongs to the family of Class IIS restriction enzymes, which recognize non-palindromic sequences, and cleave at a site that is separate from their recognition sequences. The latter trait gives Class IIS enzymes two useful properties. First, when a Class IIS enzyme recognition site is engineered at the end of a primer, the site is cleaved when digested. Second, overhangs created by Class IIS enzymes are template-derived and thus unique. This is in clear contrast to regular Class II restriction enzymes such as EcoRI, which creates an enzyme-defined overhang that will ligate to any EcoRI-digested end. The unique overhangs produced by Class IIS enzymes can be ligated only to their original partner.

It is often preferable to amplify the gene encoding the protein of interest via PCR prior to cloning into the pSUMO vector. The forward primer must contain the additional standard sequence:

5′-GGTCTCAAGGTNNN-3′ (SEQ ID NO:44) where GGTCTC is the Bsa I site and NNN is the first codon of the gene encoding the protein of interest. Additional nucleotides are required for the primer to anneal specifically with the gene of interest during the PCR amplification. The reverse primer may contain another restriction enzyme such as Xho I to allow for directional cloning of a gene into pSUMO. Bsa I can also be employed in the reverse primer to simplify cloning steps, for example, in the following primer:

• 5′-GGTCTCCTCGAGTTANNN-3′ (SEQ ID NO:45)
  The PCR product can be digested with both Xho I and Bsa I. A digestion reaction containing just the latter enzyme generates a product that would directionally ligate into the pSUMO vector between the Bsa I and Xho I sites of the MCS.
  Construction of pSUMO-Protein G Fusion E. coli Expression Vector:

The B2 IgG binding domain (9) from streptococcus G148 protein was synthesized by three synthetic oligonucleotides. The sequence of the gene is 5′-GT CTTAAGA CTA AGA GGT GGC ACG CCG GCG GTG ACC ACC TAT AAA CTG GTG ATT AAC GGC AAA ACC CTG AAA GGC GAA ACC ACC-3′. (SEQ ID NO:46) The 81 bps oligo sequence is 5′-GCC GTT ATC GTT CGC ATA CTG TTT AAA CGC TTT TTC CGC GGT TTC CGC ATC CAC CGC TTT GGT GGT TTC GCC TTT CAG-3′. (SEQ ID NO:47) The 86 pbs oligo sequence is 5′-CAG TAT GCG AAC GAT AAC GGC GTG GAT GGC GTG TGG ACC TAT GAT GAT GCG ACC AAA ACC TTT ACC GTG ACC GAA TAA GGT ACC CC-3′ (SEQ ID NO:48). The bolded nucleotides refer to the AflII and Kpn1 sites that flank the protein G domain. ACG is the first amino acid residue of the domain. The above three oligos were annealed using the Life Technologies protocol. The annealed fragments were extended by Poll enzyme. The resultant gene was PCR amplified by the following oligo primers GI forward 5′-CTT GTC TTA AGA GGT-3′ (SEQ ID NO:49) and G2 reverse primer 5′-GCT GGG TAC CTT ATT CGG TCA-3′ (SEQ ID NO:50). The above protein G gene was cloned at the AflII and Kpn1 site of the human ubiquitin gene and expressed as ubiquitin-protein G fusion protein in an E. coli pET 22 expression vector (Novagen). The protein G sequence was in turn amplified from the ubiquitin-protein G fusion plasmid by using the primers 5′-GGTCTCAAGGTACGCCGGCGGTGACCACCT-3′ (SEQ ID NO:51) and 5′-AAGCTTATTATTCGGTCACGGTAAAGGTTT-3′ (SEQ ID NO:52) and inserted in pSUMO to generate pSUMO-protein G expression vector.

Construction of E. coli SUMO-β-galactosidase Expression Vector.

E. coli β-galctosidase was amplified using pfu (Stratagene) a preparation of genomic DNA from BL21(DE3) (Stratagene) as a template and the primers 5′-GGTCTCAAGGTATGACCATGATTACGGATTCACT-3′ (SEQ ID NO:53) and 5′-AAGCTTATTATTATTATTTTTGACACCAGACC-3′ (SEQ ID NO:54). The PCR products were purified and double digested with Bsa I and Hind III. These were then ligated into the vector pET24d6×HisSUMO, which had been similarly digested.

Construction of E. coli pSUMO-Liver X Receptor (LXR) Expression Vector:

The PCR products of the LXR from amino acid residue 189 to the end of the protein that spans the ligand binding domain was digested with BsaI and HindIII and ligated into the pSUMO vector, also digested with Bsa1 and HindIII.

Construction of E. coli pSUMO-MAPKAP2 Expression Vector:

The fragment of MAPKAP2, encoded in the plasmid pMON45641, was amplified by PCR and cloned into pET24d 6HisSUMO vector by designing PCR primers that flank the sequence shown FIGS. 8A and 8B. The SUMO vector was digested with Bsa I site and Hind III. The cloning procedure yields a fusion protein, which, upon expression, purification and cleavage, generates the desired protein whose first amino acid is a glutamine (CAG).

Construction of E. coli pSUMO-Tyrosine Kinase Expression Vector:

For the tyrosine kinase, both, the SUMO fusion and unfused expression vectors were designed. As described above the region of kinase was cloned by PCR flanked with BsaI and Hind III sites that were cloned in to similarly digested pSUMO.

Construction of E. coli pSUMO-β-Glucuronidase Expression Vector:

E. coli β-glucuronidase was the kind gift of Ben Glick, University of Chicago) and amplified with the primers 5′-GGTCTCAAGGTATGCAGATCTTCGTCAAGACGTT-3′ (SEQ ID NO:55) and 5′-AAGC TTATTATTGTTTGCCTCCCTGCTGCG-3′ (SEQ ID NO:56).

Construction of E. coli SUMO-Hydrolase Expression Vector:

C-terminal His-tagged SUMO hydrolase/protease Ulp (403–621)p (21) (27) was expressed from pET24d in Rosetta (DE3) pLysS (Novagen). The recombinant protein was purified using Ni-NTA agarose (Qiagen) and buffer exchanged into 20 mM Tris-HCl pH 8.0, 150 mM NaCl and 5 mM β-mercaptoethanol using a PD-10 column (AP Biotech). About 2 ug of the pure protein was analyzed on gels and data shown in FIG. 6 lane Ulp1. The protein was almost 90% pure as judged by SDS-PAGE analysis.

Construction of E. coli UBL-GFP Fusion Vectors.

DNA sequences encoding ubiquitin (Ub), SUMO, Urm1, Hub1, Rub1, Apg8, and Apg12 were PCR-amplified using Deep-Vent polymerase (NEB) and yeast strain DNA to generate a template. Full-length human ISG15 cDNA was a kind gift of Dr. A. Haas, Medical College of Wisconsin, Milwaukee. A unique NcoI site followed by 6His sequence was introduced by PCR at the 5′-end of each Ubl cDNA. Primer sequence at the 3′-end included unique Esp3I and HindIII sites. PCR products were digested with NcoI/HindIII and inserted into respective sites of pET24d vector (Novagen) as described above. Full length GFP sequence (Clontech Cat # 60610-1) flanked by Esp3I and HindIII sites, respectively, was PCR-amplified and cloned into pCR4-TOPO-TA vector (Invitrogen). Esp3I/HindIII digested GFP-encoding gene was inserted into respective sites of pET24d-UBLl plasmids, creating final UBL-GFP expression vectors for E. coli. In toto, there were nine plasmid constructs coding for the following structures: 6His-Ubl-GFP. All plasmids were sequenced to confirm the expected structure.

Design and Construction of Yeast UBL-Fusion Vectors:

Saccharomyces cerevisiae has been used as a eukaryotic model for all the experiments involving yeast. All of the expression vectors for these studies were designed on multicopy yeast vectors that contain tryptophan or leucine as a selectable marker and 2μ as an origin of replication (22). Proteins were expressed as unfused products or as ubiquitin, SUMO or other UBL fusion proteins.
Construction of the β-Glucuronidase Yeast Expression Vectors:

To demonstrate that UBLs increase the level of secretion of the protein to the media, in addition to enhancing the level of expression, expression vectors were constructed with and without ubiquitin. We have also compared ubiquitin fusion and SUMO fusion using GFP as a model protein (see FIG. 9 and FIG. 10). pRS425-GUS plasmid was produced by cloning the XhoI-SacI fragment (containing E. coli β-Glucuronidase (GUS)) from plasmid pGUS1 (25, 22) into the XhoI-SacI sites of plasmid pRS425 (32). The next construction involved addition of a promoter, and resulted in the plasmid pRS425-ADH1p-GUS. The fragment XhoI-HindIII (containing the ADH1) was inserted into the XhoI-HindIII sites of the plasmid pRS425-GUS. The ADH1 promoter XhoI-HindIII fragment was cloned using polymerase chain reaction (PCR), amplifying the ADH1 promoter from the plasmid pGRIP1(37). The following primers were used to amplify the full length ADH1 promoter: ADH1-XhoI: 5′-gctcgagagcacagatgcttcgttg-3′ (SEQ ID NO:57), and ADH1-HindIII: 5′-gcaaagcttggagttgattgtatgc-3′ (SEQ ID NO:58). The underlining indicates the nucleotide sequence of the XhoI and HindIII restriction sites. PCR of the DNA fragment involved amplification in 30 cycles (96° C.—30 sec., 54° C.—1 min. and 72° C.—3 min.) using high replication fidelity Deep Vent Polymerase (New England Biolabs). The PCR product was then digested with XhoI and HindIII, and subsequently cloned into the XhoI-HindIII sites of pRS425-GUS. Construction of the next set of plasmids involved a change in promoter. The following two plasmids were constructed to give expression vectors containing either a methionine or proline junction between the ubiquitin and the GUS. pRS425-GPDp-Ub(Methionine)-GUS and pRS425-GPDp-Ub(Proline)-GUS were similarly constructed using both pre-constructed plasmids and PCR amplification. The final expression construct was pRS425-CUP1p-SUMO-GUS, which was the only plasmid produced with the CUP1, copper regulated promoter. This plasmid was digested with the enzymes Bg1II and NsiI, releasing the CUP1 promoter (6). The CUP1 fragment was then ligated to pRS425-GPDp-Ub-GUS, having also been digested with Bg1II-NsiI.

Construction of SUMO-N-GFP Yeast Expression Vector:

To determine what variety of N-terminal variant amino acids at the junction of SUMO and GFP can be cleaved in yeast we designed SUMO-GFP vectors in which all 20 amino acid residues were encoded at the N-terminus of GFP. Essentially all 20 SUMO-X-GFP vectors designed for E. coli expression were digested with Bsa I-Hind III, and the inserts were purified. The 20 inserts were cloned in Yep12 that was slightly modified. Specifically, YeEpSW was generated by digesting Yep12 with Bam HI and SacI. The CUP1 promoter region was recovered from the fragment by PCR. A polylinker was created at the 3′ end of CUP1 with a variety of restriction sites including NcoI and Xho1. All 20 SUMO-GFPs (N end variants) were digested with NcoI-XhoI enzymes and cloned directly YepSW. The resultant vector YepSW-SUMO-eGFP utilizes tryptophan selection and expresses SUMO-GFP proteins under the control of the copper promoter. All vectors were sequenced to ensure correct codons at the junction of SUMO and GFP.

Construction of UBL-GFP Fusion Yeast Expression Vectors:

Construction of the UBL-GFP fusion vectors for E. coli has been described above. In order to make UBL yeast expression vector NcoI/XhoI fragments carrying GFP alone and all the Ubl-GFP fusions were inserted into respective sites of pYEp SW (see above) that was similarly digested with NcoI/XhoI. Insertion of UBL-GFP cassette in Yep SW (See FIGS. 39 and 40A–40F), allows copper inducible expression of Ubl-GFP fusions in yeast system.

Design and Construction of Recombinant Baculovirus for SUMO and Ubiquitin GFP Fusion Expression:

To demonstrate that attachment of SUMO or ubiquitin to GFP increases its expression and enhances secretion into the media, several GFP fusion vectors were designed with different configurations of gp67 secretory signals. The basic GFP vector for expression is essentially based on E. coli vectors described above. Derivatives of this vector representing each candidate gene have been constructed by designing PCR primers. The construction of GFP plasmid transfer vectors for baculovirus is described. To help appreciate the rationale for the secretory signal in the context of GFP-fusion, see the diagrammatic representation shown in FIG. 11. Single letter code refers to unfused GFP (E); gp67-sec signal-GFP (G); ubiquitin-GFP (U); SUMO-GFP (S); gp67-Ub-GFP (GU); Ub-gp67-GFP (UG); gp67-SUMO-GFP (GS); and SUMO-gp67-GFP (SG).

(i) pFastbacE. A synthetic oligonucleotide containing the Esp3I site was inserted between BamHI and EcoRI cloning site of the transfer vector pFastbac1, which had been modified by removing Esp3I site from Gmr region. (ii) pFastbacG. The signal sequence of the gp67 gene derived from pACSecG2T was isolated by PCR using 2 primers (f-gp67 and r-gp67), digested with Bg1II and EcoRI in the next step, and then inserted between BamHI and EcoRI cloning sites of the transfer vector pFastbacE. (iii) pFastbacS. A full-length SUMO gene derived from pET SUMO was generated by PCR using 2 primers (f-bacsmt and r-bacsmt), digested with BsaI and EcoRI in the next step, and then inserted between BamHI and EcoRI cloning sites of the transfer vector pFastbacE. (iv) pFastbacG/S. The signal sequence of the gp67 gene in the pACSecG2T vector was generated by PCR using 2 primers (f-fusgp67 and r-fusgp67), and inserted between BamHI and EcoRI cloning sites of the transfer vector pFastbacE to create a new pFastbacG, which was used for fusion with SUMO afterward. A full-length SUMO gene derived from pET SUMO as described above (iii) was digested with BsaI and SacI and inserted between Esp3I and SacI cloning sites of the new transfer vector pFastbacG. (v) pFastbacS/G. A full-length SUMO gene derived from pET SUMO was generated by PCR using 2 primers (f-fussmt3 and r-fusgp67) and inserted between BamHI and EcoRI cloning sites of the transfer vector pFastbacE to create the new pFastbacS, used for fusion with gp67 afterward. The signal sequence of the gp67 gene derived from pACSecG2T as described above (ii) was digested with BsaI and SacI, and then inserted between the Esp3I and SacI cloning sites of the new transfer vector pFastbacS.

Preparation of Baculovirus Stocks and Cell Growth.

Transfer vector constructs based on the pFastbac 1 shuttle plasmid (Invitrogen, Inc.) were transposed in DH10Bac E. coli competent cells to transfer the respective e-GFP fusion sequences into recombinant virus DNA by site-specific integration. After alkaline lysis of transformed (white colonies) of E. coli cells, which contain recombinant virus (bacmid) DNA, and extraction of the recombinant bacmid DNA, the bacmid DNA was used to transfect Spodoptera frugiperda (Sf9) insect cells, in which virus replication occurs. The virus was then amplified to produce passage 2 (for long-term storage) and passage 3 virus (for working) stocks by infection of fresh Sf9 cell cultures and used directly to infect cells for fusion protein expression. Virus infectivity (pfu/ml) was determined by titration in Sf9 cells using the BacPAK™ Rapid Titer Kit (BD Sciences Clontech, Inc.). A 50 ml culture of Hi-Five cells at concentration of 1×106 cells/ml, was infected with recombinant virus at MOI=5 in Express Five media (serum free media). The cells were grown in 100 ml spinner flask at 27° C. Every 24 hours, cell viability was determined by trypan blue and cell counting. 5 ml of the suspension culture was removed at 24 hour intervals, centrifuged at 500×g at 4° C. in 10 minutes. The supernatant was transferred into a fresh tube to monitor any protein that may have been secreted into the media (see below).

Analysis of Proteins from Insect Cell Compartments:

Cell pellets (from above step) were gently washed in 1 ml PBS and recentrifuged at 500×g at 4° C. for 10 minutes. All supernatant and pellets are stored at −80° C. The presence of recombinant protein in cells and media was ascertained by SDS-PAGE and Western blotting of supernatant and cell pellets. The total intracellular protein was extracted by M-PER extraction buffer (Pierce), a neutral buffer for protein extraction. The cell pellet was mixed with rapid pipetting and incubated for 1 hour on an orbital shaker. The suspension was centrifuged at 500×g at 4° C. for 10 minutes to remove debris. The supernatant contained extracted cellular proteins that were either analyzed by PAGE or stored at −80° C. To analyze the proteins present in the media, the following procedure was adopted. Trichloroacetic acid was added to 5 ml media to a final concentration of 20%. The suspension was mixed well and left on ice for three hours, and then centrifuged 500×g at 4° C. for 10 minutes. The white pellet was washed with 80% ethyl alcohol twice, and then dried. The pellet was suspended in 1 ml of M-PER buffer for PAGE to compare the distribution of control (unfused) and SUMO-fused proteins inside and outside the cell.

Methods for Analysis of Yeast Expressed Fusion Proteins:

Yeast cultures were grown in synthetic or rich media. Standard yeast and E. coli media were prepared as described (31). The yeast strain Y4727: Mata his3-Δ200 leu2-Δ0 lys2-Δ0 met5-Δ0 trp1-Δ63 ura3-Δ0 was used as a host (gift from Dr. Jeff Boeke) or BJ 1991. Yeast transformation was performed according to published procedures (8). Yeast transformants with autonomously replicating plasmids were maintained in yeast selective media. The E. coli β-Galactosidase and β-Glucuronidase proteins were expressed under the regulation of either the alcohol dehydrogenase (ADH), or Glyceraldehyde-Phosphate-Dehydrogenase (GPD) promoter or copper metallothioneine (CUP1) promoter in 2 μm multicopy plasmids with the LEU2 selective marker.

Yeast cells were transformed with appropriate expression vectors, and single colonies were grown in synthetic media minus the selectable marker. For each protein, at least two single colonies were independently analyzed for protein expression. Cells were grown in 5 ml culture overnight and, in the morning, the culture was diluted to an O.D. at 600 nm of 0.5. If the gene was under the control of copper inducible promoter, copper sulfate was added to 100 μM and the culture was allowed to grow for at least three hours. Cells were pelleted at 2000×g for 5 minutes, washed with 10 mM Tris-EDTA buffer pH 7.5. If enzymatic assays were performed, cells were disrupted in assay buffer with glass beads, 2× times the volume of the pellet. Cells were centrifuged and the supernatant was recovered for enzymatic or protein analysis. Alternatively, if the level and the type of protein was analyzed by SDS-PAGE, cell pellet was suspended in SDS-PAGE buffer and boiled for 5 mins. The suspension was centrifuged, and 10–20 ul aliquots were run on 12% SDS-PAGE.

Measurement of β-GUS Activity from Yeast:

β-Glucuronidase (GUS) is a 65 kDa protein that is a useful marker for protein trafficking. We have used GUS to determine the role of N-terminal ubiquitin on secretion of GUS in yeast. Yeast cells were transformed with various GUS vectors, grown overnight in selective liquid media at 30° C., and diluted in the liquid selective media to 0.1 OD600 (OD culture). Yeast cells were incubated in the presence of inducer in shaker at 30° C. After 4 hours of incubation, 100 μl of 2× “Z” Sarcosine-ONPG buffer (120 mM Na2HPO4, 80 mM NaH2PO4, 20 mM KCl, 2 mM MgSO4, 100 mM β-mercaptoethanol, pH 7.0, 0.4% lauroyl sarcosine) was added. (The 2× “Z” Sarcosine-buffer is freshly prepared or stored at −20° C. prior use.) We used a fluorometric assay with 4-methylumbelliferyl β-D-glucuronide as the substrate for β-GUS assay. After incubation at 37° C. for 1 hour (t incubation), the reaction was stopped by adding 100 μl of quenching solution, 0.5 M Na₂CO₃. The GUS activity was determined by reading the plates in a fluorometric plate reader. For calorimetric reactions, relative activity was calculated as following: (1000×OD reaction)/(t incubation×OD culture).

E. coli Growth, Compartmentalization and Protein Expression.

Protein expression studies were carried out in the Rosetta bacterial strain (Novagen). This strain is derived from the lambda DE3 lysogen strain and carries a chromosomal copy of the IPTG inducible T7 RNA polymerase along with tRNAs on a pACYC based plasmid. Cultures were grown in LB as well as minimal media and at growth temperatures of 37° C. and 20° C. with 100 ug/mL ampicillin and 30 ug/mL chloramphenicol. The culture was diluted 50 fold and grown to mid log (OD at 600 nm=0.5–0.7), at which time the culture was induced with 1 mM IPTG. Induction was allowed to proceed for 4–5 hrs. Upon completion of induction, cells were centrifuged and resuspended in a buffer containing 20% sucrose. To analyze protein induction in total cells, SDS-PAGE buffer was added and the protein was analyzed following SDS-PAGE and staining with Coomassie blue.

Separation of Soluble and Insoluble Fractions.

E. coli were harvested by mild centrifugation and washed once with PBS buffer. Cells were resuspended in 4 ml of PBS and ruptured by several pulses of sonication. Unbroken cells were removed by mild centrifugation (5 min at 1500×g) and supernatants were sonicated again to ensure complete cell lysis. An aliquot (5 μl) was mixed with 2% SDS to ensure that no viscosity is detected owing to lysis of unbroken cells. After ensuring that no unbroken cells remained in the lysate, insoluble material consisting of cell walls, inclusion bodies and membrane fragments was sedimented by centrifugation (18,000×g for 10 min). The supernatant was considered “Soluble fraction”.

The pellets were washed from any remaining soluble proteins, lipids and peptidoglycan as follows. Pellets were resuspended in 600 μl of PBS and to the suspensions 600 μl of solution containing 3 M urea and 1% Tri ton X100 was added. The suspension was briefly vortexed and insoluble material was collected by centrifugation as above. The PBS/Urea/Triton wash was repeated two more times to ensure complete removal of soluble proteins. The washed pellets, designated as “insoluble fraction,” consisted primarily of inclusion bodies formed by over expressed proteins. Approximately 10 μg of protein from each fraction was resolved on 12% SDS-PAGE minigels and stained with Coomassie Brilliant Blue.

Fluorescence (GFP Activity) Assessment.

GFP fluorescence was measured in soluble fractions (approx. 0.1 mg of soluble protein in a final volume of 40 μl) using Fluoroscan Accent FL fluorometer (LabSystems) with Excitation 485 nm/Emission 510 nm filter set with the exposure set to 40 sec. The data are presented in Arbitrary Units (AU).

Western Blotting.

Twenty μg of total yeast protein per lane were resolved on 12% SDS-PAGE minigel and electro-blotted to nitrocellulose membranes by standard methods. Membranes were blocked with 5% milk in TTBS buffer and incubated with rabbit anti-GFP antibodies (Clontech, cat no. 8367) at 1:100 dilution overnight at 4° C. Secondary HRP-conjugated antibodies were from Amersham. Identical gels were run in parallel and stained with Coomassie to ensure equal loading of the samples.

The various 6HisxSUMO-GFP (16) fusions were expressed in Rosetta (DE3) pLysS (Novagen) using the procedures recommended by the manufacturer. Expression levels in the absence and presence of the fusion proteins was compared by SDS-PAGE analysis. The recombinant proteins were purified using Ni-NTA agarose; (Qiagen) using procedures recommended by the manufacturer.

Cleavage of Proteins

For studies in E. coli, an organism that does not possess SUMO or ubiquitin cleaving enzymes, each cleavage reaction contained 100 ul of purified fusion protein, 99 ul of the buffer 20 mM Tris-HCl pH 8.0, 150 mM NaCl, 5 mM β-mercaptoethanol, and 1 ul of enzyme. The reactions were incubated for 3 hours at 30° C., and stopped by addition of 6× Laemmli SDS-page loading buffer followed by boiling at 95° C. for 5 minutes. The products of the cleavage reaction were analyzed by SDS-PAGE.

The following examples are provided to illustrate various embodiments of the present invention. They are not intended to limit the invention in any way.

EXAMPLE I

Attachment of C-Terminus of UBLs to N-Terminus of GFP Enhances the Expression and Solubility of the Protein in E. coli.

The design and construction of all the UBL E. coli expression vectors has been described above. The DNA sequences, accession numbers of the UBL-GFP fusion proteins, and translation frames are shown FIGS. 25–32. FIG. 4A shows the 37° C. expression pattern of GFP, Ub-GFP, SUMO-GFP, Urm1-GFP, Hub1-GFP, Rub1-GFP, Apg8-GFP, Apg12-GFP, ISG15-GFP. Un-fused GFP is generally poorly expressed in E. coli. The data show that all of the UBLs enhance the expression level of GFP to varying degrees. However, the greatest amount of induction was observed with Ub, SUMO, Urm1, Apg8 and Apg12. Induced cells were broken by sonication and soluble proteins were analyzed on SDS-polyacrylamide gels. The stained gel shows (FIG. 4A, Soluble Panel) that ubiquitin, SUMO, Urm1, Hub1 and ISG15 were able to solublize the GFP while Rub1, Apg8 and Apg12 fusion proteins were not soluble, however, fusion to these proteins did enhance the level of expression several fold. To determine if the fusion proteins were folded correctly, we determined the fluorescence properties of proteins in the soluble fraction. FIG. 4A also shows GFP fluorescence in approximately 0.1 mg of soluble protein in a final volume of 40 ul using Fluoroscan Accent FL fluorometer (LabSystems) with Excitation 485 nm/Emission 510 nm filter set with the exposure set to 40 sec. The data are presented in Arbitrary Units (AU) and show that Ub, SUMO, Urm1, Hub1 and ISG15 produced GFP protein that was able to fluoresce and, thus, was folded correctly. Fusions of GFP with Rub1, Apg8 and Apg12 were induced in large amounts but were not soluble and did not show any fluorescence.

In addition, it is shown that ISG15 plays a role in immune response (24). Thus presentation of ISG15 as a fusion protein is a viable tool for novel vaccine candidates. Similarly, Apg8 and Apg12 translocate protein to compartments in the cell for autophagy (30).

Similar experiments were performed with all the UBL-GFP fusion proteins, but the induction was performed at 26° C. overnight. The data shown in FIG. 4B confirms the finding in FIG. 4A. Almost all of the UBLs except Hub1 showed dramatically enhanced expression of GFP after fusion. In the case of SUMO, the level of expression was increased about 20 fold. Analysis of soluble fraction showed that Ub, SUMO, Urm and ISG15 were able to solubilize fused GFP (see FIG. 4B, Soluble panel). Functional analysis of fusion GFP was performed by fluorescence from the soluble fraction. This data confirms the observation made in FIG. 4A. Combining all the data from the induction studies demonstrates that fusion of all the UBLs to GFP enhances expression level from 2–40 fold. In addition, Ub, SUMO, Urm1, Hub1 and ISG15 also increase the solubility of the GFP. These UBLs are therefore capable of producing correctly folded proteins in E. coli.

To gain more insight into the role of UBLs in enhancement of expression and solubility, we have tested the SUMO-fusion systems with other proteins as well. Serine threonine kinases, tyrosine kinase and human nuclear receptor have proven difficult to express in E. coli. Researchers have opted to use tissue culture systems to express soluble kinases of receptors. FIG. 5 shows expression 6His-SUMO-Tyr-Kinase and unfused Tyr-Kinase in E. coli using LB or minimal medium (MM), and purified on Ni-NTA resin as described previously. The small fraction of resin was boiled with 1×SDS-PAGE sample buffer and aliquots were resolved on the 12% SDS-PAGE. Equal amounts of E. coli culture were taken for SUMO-Tyr-kinase and unfused Tyr-kinase and purification was performed under identical conditions. The stained gel in FIG. 5 shows that SUMO fusion increases the yield of the kinase at least 20 fold, in cells grown in LB media. FIG. 6 also shows the pattern of the SUMO-Try kinase that was eluted from Ni-NTA by 100 mM EDTA or 250 mM imidazole. These data further demonstrate that SUMO fusion enhances the expression of difficult to express protein such as Tyr-kinase, and that the expressed fusion protein is soluble.

Human nuclear receptor proteins, such as steroid receptors, contain ligand-binding domains. These proteins have proven hard to express in soluble form in E. coli. We have used human liver X receptor (LXR) ligand binding domain to demonstrate that SUMO fusion promotes solubility of the protein in E. coli. The ligand-binding domain of LXR was expressed as SUMO fusion in Rosetta plysS cell at 20° C. or 37° C. and the pattern of soluble and insoluble protein was analyzed. FIG. 7 shows the stained SDS-polyacrylamide gel demonstrating that about 40% of the LXR protein was solublized by SUMO fusion, see lane CS in 20° C. box in FIG. 7 (predominant band in 40 kDa range). If the cells were induced at 37° C., hardly any SUMO-LXR was soluble although the level of protein induction had increased dramatically. Further proof that SUMO promotes solubility of previously insoluble proteins was gained by expressing MAPKAP2 kinase as a SUMO-fusion in E. coli. FIGS. 8A and 8B shows induction kinetics in E. coli cells expressing kinase at 20° C. and 37° C. Numbers at the top of the gel, 0.1, 0.25 and 0.5 refer to the mM concentration of inducer IPTG, in the culture. The original induced culture (I), supernatant from lysed cells (S) and resuspended pellet (P) were analyzed on 12% SDS-PAGE. The data clearly demonstrate that 90% of the SUMO kinase is soluble when the cells are induced at 20° C. with 0.25 mM IPTG. Although induction at 37° C. allows greater degree of expression, more than 50% of the kinase is still insoluble under these conditions. Cleavage of SUMO-MAPKKAP2 kinase by SUMO hydrolase is described in Example III. Also see FIG. 18.

Overall, these results show that in bacteria, fusion of UBLs to GFP increases the level of expression from 2–40 fold. Some of the UBLs such as Ub, SUMO, Urm1, Hub1, and ISG15 solublize otherwise insoluble proteins. In particular, SUMO has been demonstrated to increase solubility of kinases and LXR α under controlled temperature induction from 50–95% of the total expressed protein.

EXAMPLE II

Sumo-Fusion Expression in Yeast and Insect Cells

Fusions of C-Terminal UBLs to the N-Terminus of GFPs are Cleaved in Yeast

To further assess the utility of UBL fusion in eukaryotic cells we expressed all of the UBL-GFP fusions previously described in FIG. 4 in yeast. S. cerevisiae BJ1991 strain was transformed with either YEp-GFP or YEp-UBL-GFP fusion constructs using standard procedures. Positive clones were grown in YPD medium and induced with 100 μM CuSO₄ at cell density OD600=0.2 for 3.5 hours. Total cell extracts were prepared by boiling the yeast cells in SDS-PAGE buffer. Twenty ug of proteins were analyzed on 12% SDS gels. A replica gel was stained in Coomassie blue and another gel was blotted and probed with antibodies against GFP. Data in FIG. 9 shows that Ub-GFP, SUMO-GFP and ISG15-GFP fusions were efficiently cleaved in yeast, while Rub1-GFP fusion was partially cleaved. Apg8-GFP fusion was cleaved into two fragments. It is noteworthy that all the UBL-GFP fusions were designed with methionine as the first amino terminus. GFP fusion with Urm1, Hub1 and Apg12 expressed well, but were not cleaved in yeast. There was a modest increase in expression of GFP following fusion with Ub, SUMO, ISG15 and cleavage in yeast. Generally we have observed 10–20 fold increase in the level of protein expression following fusion to UBL in prokaryotes and eukaryotes (see FIGS. 4B, 10 and 11). The reason for the modest increase in GFP fusion following cleavage is that the cells were grown in induction media containing 100 μM copper sulfate in rich YPD media. Rich media contains many copper binding sites, and less free copper is available to induce the gene. A nearly 100-fold increase in GFP production has been observed with a variety of N-terminal fusions when cells were induced with 100 μM copper sulfate in synthetic media. See FIG. 10.

Generation of New Amino Termini:

The identity of the N-terminus of a protein has been proposed to control its half-life (the N-end Rule) (35). Many important biopharmaceuticals such as growth factors, chemokines, and other cellular proteins, require desired N-termini for therapeutic activity. It has not been possible to generate desired N-termini, as nature initiates translation from methionine, but the SUMO system offers a novel way to accomplish this.

To demonstrate that all N-termini of GFP in SUMO-GFP fusions were efficiently cleaved when expressed in yeast, a comprehensive study of SUMO-GFP with 20 N-termini was carried out. Multi-copy yeast expression plasmids were designed as described above. Plasmids were transformed in yeast strain BJ 1991, four single colonies were selected, and the levels and cleavage patterns of two of the strains were analyzed by SDS-PAGE and western blotting. Data from Western blots of a single colony is presented in FIG. 10. These results are in agreement with our in vitro studies of purified SUMO-X-GFPs (from E. coli) and its cleavage pattern of SUMO hydrolase. All of the SUMO-GFP fusions were cleaved efficiently except those containing proline at the junction (see FIG. 10, middle panel lane “Pro”). It is also interesting to note that SUMO-Ileu-GFP was partially cleaved during the phase of copper induction. All of the genes are under the control of copper inducible promoter. It is possible that SUMO-Ileu-GFP is resistant to cleavage due to the non-polar nature of the residue at the +1 active site of SUMO hydrolase. In this respect SUMO-Val-GFP was also partially resistant to cleavage in vivo (see lower most panel lane labeled “Val”). It is clear from these results that SUMO-Pro-GFP fusion was completely resistant to cleavage by yeast SUMO hydrolases as no GFP was observed (see lane “pro” in middle panel of FIG. 10). This data is consistent with our previous observations. See FIG. 15. Another important aspect of these findings is that fusion of SUMO with various N-termini of GFP appears to increase the expression of almost all the proteins, although to various degrees. For example Cys-GFP, Gly-GFP and His-GFP accumulated in greater amounts as compared to other N-terminal GFPs. A direct comparison of the increase in the level of GFP following fusion to SUMO can be made by comparing the level of un-fused GFP (see last lanes of lower most panel in FIG. 10). Although 20 ug of yeast proteins were loaded on SDS-PAGE the GFP signal was not detected. To ensure that we were not dealing with mutation or any artifact, we loaded a protein sample from another single colony that was induced in under similar conditions and the sample was loaded next to the previous GFP. No signal was detected, suggesting that unfused GFP is made in very small amounts that cannot be detected under the present experimental conditions, (i.e., a four hour induction with copper sulfate). These studies show that fusion with SUMO leads to a dramatic increase in the amount of protein expressed in yeast. All of the N-terminal fusions are cleaved by endogenous SUMO hydrolases except when the N-terminal residue is proline. Thus for enhanced expression of a protein in eukaryotes permanent attachment of SUMO is not required as significant (˜100 fold) increased accumulation of the protein was observed even after the cleavage of SUMO. At the same time, SUMO-pro-fusions are also useful as 6×His-SUMO can be used to purify the protein from yeast, and the SUMO moiety can be removed with 10 times greater amounts of the SUMO hydrolase (see example III).

Previous studies have shown that attachment of ubiquitin to the N-termini of proteins in yeast enhances expression, and protein fusions containing all amino acid at the N-terminal residue, except proline, are efficiently cleaved in yeast (2, 10, 34). However, these technologies have several drawbacks. Firstly, none of the deubiquitinating enzymes (DUBs) have been shown to efficiently cleave ubiquitin fusion proteins of varying sizes and structures (3,1), despite the fact that they were discovered more than 15 years ago (35, 19, 3). Secondly, and perhaps more importantly, ubiquitin predominantly functions as a signal for proteolysis (14). Therefore, for physiological reasons and for the lack of robust cleavage of artificial ubiquitin-fusions by DUBs, the ubiquitin gene fusion system has not been successfully developed for commercial applications. We have observed that the SUMO system appears to perform in a manner that is remarkably superior to that of ubiquitin, as SUMO and other UBL fusions enhance protein expression and solubility in prokaryotes. In addition, many of the UBLs increase expression of GFP, following the cleavage of UBL in yeast. Unlike the ubiquitin-fusion system, which may direct the protein to the ubiquitin proteosome pathway, the current cleavage of fusion-protein in yeast is the result of C-terminal fusion with SUMO, and proteins generated with novel N-termini are not subject to degradation by the ubiquitin-proteosome pathway. This is one of the reasons that large amount of GFP has accumulated in yeast after cleavage of the SUMO fusion (see FIG. 10).

N-Terminal Attachment of Ubiquitin Promotes Protein Secretion:

To date, a role for ubiquitin in the secretion of proteins has not been determined. We have assessed whether N-terminal fusion of ubiquitin to a protein promotes its secretion in yeast. Several yeast expression vectors that express E. coli β-glucoronidase (GUS) were designed. All of the yeast GUS expression vectors described in Table 2 are engineered under the control of the strong glycolytic GPD promoter that expresses constitutively. Some of the constructs were also expressed under the control of a copper regulated metallothionein promoter (CUP1) as well. CUP1 promoter driven synthesis of the SUMO-GUS constructs was induced by addition of 100 μM copper sulfate and incubation of 3 hours. To determine the level of GUS from media, cells were harvested by centrifugation at 2000×g for 10 mins. Supernatant was collected and equal amounts of aliquots were assayed for enzymatic activity or western blot analysis as described above. For the comparative study, all strains were treated identically and grown at the same time to equal O.D, and the assays were performed at the same time. To examine intracellular enzymatic activity, the cells were harvested by centrifugation and washed with Tris EDTA buffer, pH 7.5. The cell pellets were suspended in sarcosine buffer and ruptured with glass beads at 4° C., three times by vigorously vortexing. Supernatant was collected for assay of the enzymatic activity. The amount of protein secretion was determined by estimating relative activity of the enzyme in the media. The data is shown in Table 2.

TABLE 2
Ubiquitin-GUS Expression and Secretion in Yeast
GUS activity was measured as described. It was not possible to measure
specific units of GUS in the media as yeast grown in synthetic media.
Yeast secretes little protein and current methods of protein estimation,
BioRad kit cannot estimate the protein, the data was presented as +
where one + is equal to 2 units of GUS as described in invention.
− Sign means no GUS activity was detected.
			GUS	GUS
			Activity	Activity
Vector		Signal	Inside	In
(pRS425)	Promoter	Sequence	Cell	Supernatant
ADH1-GUS1	ADH1	—	+++	−
GPD-α-factor-	GPD	α-factor	++
GUS1
GPD-Ub-GUS1	GPD	Ubiquitin	++++	++++
GPD-Ub-α-factor-	GPD	Ubiquitin-α-	++++	−
GUS1		factor
GPD-α-factor-Ub	GPD	α-factor-	++	−
(pro)-GUS1		Ubiquitin(pro)
GPD-α-factor-Ub	GPD	α-factor-	++	−
(met)-GUS1		Ubiquitin(met)
CUP1-Ub-GUS1	CUP1	Ubiquitin	++++	++

The following conclusions are drawn from this study.

• 1) Fusion of ubiquitin to GUS leads to a several fold increase when yeast extracts were analyzed by enzymatic assays.
• 2) Insertion of proline at the junction of ubiquitin and GUS did not allow cleavage of the ubiquitin-GUS fusion protein.
• 3) The attachment of alpha factor secretory sequences to the N-terminus of ubiquitin-fusion did not have show any appreciable increase in secretion of the protein into the media.
• 4) Presence of alpha factor sequences between ubiquitin and GUS did not lead to any increase in extracellular level of GUS activity.
• 5) Greatest amount of secretion was observed with ubiquitin-Met-GUS. These observations suggest that endogenous secretory sequences of GUS in the context of ubiquitin promote the best secretion for GUS. To this end the current data from yeast correlates very well with the ubiquitin-GFP protein secretion in insect cells (see FIG. 13).
  Fusion of SUMO and Ubiquitin to the N-Terminus of GFP Promotes Enhanced Expression and Secretion in Insect Cells.

The role of SUMO in enhanced expression and secretion of proteins in cultured cells has also been studied in insect cells. Baculovirus vectors expressing SUMO-GFP constructs and appropriate controls have been described above. See FIG. 11A for the orientation gp67 secretory signals in the SUMO-GFP constructs. Data from a 24 hour infection is shown in FIG. 12. Panel A shows intracellular protein analysis by Western blots. It is clear that fusion with ubiquitin and SUMO promotes a large increase in the amount of protein (compare lane E with lane U and S). Insertion of gp67 signal sequences to the N-terminus of SUMO leads to further increase in the amount of protein in insect cells (compare unfused GFP lane E with gp67-SUMO-GFP lane GS). On the other hand attachment of gp67 signal sequence to the N-terminus of GFP (lane G, UG or SG) did not increase the level of protein expression, to the contrary there was diminution of signal when gp67 was attached to N-terminus of GFP (lane G) or between SUMO and GFP (lane SG). We estimate that in the level of expression in the context of gp67-SUMO-GFP is 20× fold higher as compared to unfused GFP (lane E) or 40× fold higher as compared to gp67-GFP (lane G). No unfused GFP was secreted by any of the constructs at 24 hour post infection, as shown in blot in FIG. 12 panel B. These results show that fusion with SUMO leads to a dramatic increase in expression of GFP in insect cells. Additionally, both SUMO-GFP and gp67-SUMO-GFP were efficiently cleaved by endogenous SUMO hydrolases.

Similar experiments were performed with cells 48 hours post infection. The data in FIGS. 13A and B show that the pattern of intracellular expression was similar to the one seen in 24 hours of infection; however, large amounts of ubiquitin and SUMO-GFP protein were secreted at 48 hour post infection. Examination of the blots from media and intracellular protein show that reasonable expression of unfused GFP was observed inside the cell, but hardly any protein was secreted in the media (compare lane E of panel A and panel B in FIG. 13). Attachment of gp67 to the N-terminus of SUMO-GFP leads to the greatest amount of protein secreted into the media (see lane GS in panel B). Another important finding is that attachment of ubiquitin without any signal sequences shows very high secretion of GFP in the media. This result is completely consistent with our finding that attachment of ubiquitin to the N-terminus of GUS promotes the greatest amount of secretion of GUS into the yeast media.

We have also discovered that SUMO-Pro-GFP fusion was not cleaved by endogenous SUMO hydrolases in insect cells (FIG. 13C). Although some non-specific degradation of SUMO-Pro-GFP was observed in these experiments (see lane S-P in FIG. 13C), we conclude that unlike SUMO-GFP, SUMO-Pro-GFP is not cleaved in insect cells. This observation is also consistent with the finding in yeast that SUMO-Pro-GFP is not cleaved in cells while other N-terminal GFP fusions are processed in yeast.

Further confirmation of these observations was obtained by fluorescence imaging of the cells expressing GFP fusion proteins. FIG. 14 shows that cells expressing GFP and fusion GFP fluoresce intensely. The fluorescence imaging was the strongest and most widely diffused in cell expressing gp67-SUMO-GFP and Ub-GFP. These cells show the largest amount of GFP secreted into the media (FIG. 13 panel B). It appears that secretory signal attachment directly the to N-terminus of GFP produces less GFP in the media and inside the cells. This observation is borne out by low fluorescence intensity and granulated pigmented fluorescence (see panel G-eGFP, S/G-eGFP and U/G-eGFP). These data have led to the following conclusions:

• 1) The increase in the amount of SUMO-fusion protein expression in insect cells was several-fold higher (20–40 fold) than that of unfused protein, as determined by and Western blot analysis.
• 2) All of the SUMO-GFP constructs that contain methionine at the +1 position were cleaved except SUMO-Proline-GUS. This aspect of the SUMO-fusion technology allows us to express proteins that are stably sumoylated.
• 3) Attachment of ubiquitin to the N-terminus of GFP led to dramatic enhancement in secretion of the protein in the media. Ubiquitin promotes secretion of proteins that may or may not have endogenous secretory signal. Thus, N-terminal ubiquitination may be utilized as a tool to enhance secretetion of proteins in eukaryotic cells.
• 4) N-terminal SUMO also promotes secretion of protein in insect cells.

EXAMPLE III

SUMO Protease ULP1 Cleaves a Variety of SUMO-Fusion Proteins:

Properties and Applications in Protein and Peptide Expression and Purification

Yeast cells contain two SUMO proteases, Ulp1 and Ulp2, which cleave sumoylated proteins in the cell. At least eight SUMO hydrolases have been identified in mammalian systems. The yeast SUMO hydrolase Ulp1 catalyzes two reactions. It processes full length SUMO into its mature form and it also de-conjugates SUMO from side chain lysines of target proteins. Examples I and II establish our findings that attachment of SUMO to the N-terminus of under-expressed proteins dramatically enhances their expression in E. coli, yeast and insect cells. To broaden the application of SUMO fusion technology as a tool for expression of proteins and peptides of different sizes and structures, the ability of Ulp1 to cleave a variety of proteins and peptides has been examined. Purified recombinant SUMO-GFPs were efficiently cleaved when any amino acid except Proline is present in the +1 position of the cleavage site. Similar properties of SUMO hydrolase Ulp1 were observed when Sumo-tyrosine kinase, Sumo-protein G, Sumo-β-GUS, and SUMO MAPKAP2 kinase were used as substrates. The in vitro activity of the enzyme showed that it was active under broad ranges of pH, temperature, and salt and imidazole concentration. These findings suggest that the Ulp1 is much more robust in cleavage of the SUMO-fusion proteins as compared to its counterpart, ubiquitin-fusion hydrolase. Broad specificity and highly efficient cleavage properties of the Ulp1 indicate that SUMO-fusion technology can be used as a universal tag to purify a variety of proteins and peptides, which are readily cleaved to render highly pure proteins.

The following materials and methods are provided to facilitate the practice of Example III.

Affinity Purification and Cleavage of SUMO Fusion Proteins with SUMO Hydrolase.

The following table lists the solutions required for the affinity purification and cleavage procedures:

Solution       Components

Lysis buffer   25 mM Tris pH 8.0; 50 mM NaCl
Wash Buffer    25 mM imidazole; 50 mM Tris pH 8.0; 250 mM NaCl;
               (optional) 5–10 mM β-mercaptoethanol (protein
               dependent)
Elution Buffer 300 mM imidazole; 50 mM Tris pH 8.0; 250 mM
               NaCl; (optional) 5–10 mM β-mercaptoethanol (protein
               dependent)
SUMO hydrolase 50 mM Tris pH 8.0; 250 mM NaCl; 5 mM β-
(Ulp1)         mercaptoethanol (protein dependent)
Cleavage Buffer

From typical 250 ml cultures, the samples are pelleted by centrifugation, and supernatants are removed by decanting. Generally, from 250 ml of culture, 1.0–1.5 grams of wet cells are produced. Pelleted cells are then resuspended in 5–10 ml of lysis buffer. RNase and DNase are added to final concentration of 10 ug/ml lysis solution. Samples are kept on ice throughout the sonication procedure. Using an appropriate tip, the samples are sonicated 3–5 times for 10 second pulses at 50% duty cycle. Sonicates are incubated on ice for 30 minutes; if the samples are viscous after this time, the sonication procedure is repeated. Lysed samples (in lysis solution) are loaded onto 1-ml columns. The columns are washed with 5 to 10 volumes of wash buffer (wash fractions are saved until the procedure is complete). Columns are developed with 2.5 ml of elution buffer, and SUMO hydrolase cleavage is performed by one of two methods: 1) cleavage is performed in elution buffer, with SUMO hydrolase added at 50 ul/250 ml buffer, samples incubated at room temperature for 2 hr or overnight at 4° C., and cleavage monitored by gel electrophoresis; 2) imidazole is first removed by dialysis, gel filtration, or desalting, samples are then resuspended in SUMO hydrolase cleavage buffer, SUMO hydrolase is added at 50 ∥l/2.5 ml buffer, and samples are incubated at room temperature for 2 hr or at 4° C. overnight, with cleavage monitored by gel electrophoresis. Units of SUMO hydrolase are defined as the amount of enzyme that cleaves 1 ug of pure SUMO-Met-GFP (up to 95%) in 50 mM Tris-HCl pH 8.0, 0.5 mM DTT, 150 mM NaCl at room temperature in 60 minutes.

After cleavage, protein can be stored at 4° C., or subjected to purification.

##STR00001##

The expression and purification of carboxy terminus of Ulp1p is described above.

In Vitro Cleavage Experiments

The various His6smt3XeGFP fusions were expressed in Rosetta (DE3) pLysS (Novagen). The recombinant proteins were purified using Ni-NTA agarose (Qiagen). The comparative in vitro cleavage reactions were carried out by first normalizing the amount of the various fusions in each reaction. This was done by measuring the fluorescence properties of the purified fusion proteins using the fluorimeter Fluoriskan II (Lab Systems) and then diluting the more concentrated samples with the Ni-NTA agarose elution buffer (20 mM Tris-HCl pH 8.0, 150 mM NaCl 300 mM Imidazole and 5 mM beta-mercaptoethanol), such that their fluorescence values equaled that of the lowest yielder. Each cleavage reaction contained 100 ul of protein, 99 ul of the buffer 20 mM Tris-HCl pH 8.0, 150 mM NaCl and 5 mM beta-mercaptoethanol and 1 ul of enzyme. The reactions were incubated for 3 hours at 30° C. after which they were stopped by addition of 6× Laemmli SDS-page loading buffer followed by boiling at 95° C. for 5 minutes. The products of the cleavage reaction were analyzed by SDS-PAGE.

Proline cleavage experiments were carried out in a fashion similar to those described above. The purified His6smt3PeGFP was buffer exchanged into 20 mM Tris-HCl pH 8.0, 150 mM NaCl and 5 mM beta-mercaptoethanol using a PD-10 column. A 10 fold increase in the amount of Ulp1 were added to each reaction. Digestions were incubated for 3 hours at 30° C. All reactions were stopped by addition of Laemmli loading buffer and analyzed by SDS-page.

FIG. 15 shows the stained SDS-PAGE analysis of all the SUMO-X-GFPs and their digestion by SUMO hydrolase. The findings clearly show that Ulp1 hydrolase was able to cleave all the SUMO-GFP fusions except proline. These finding are similar to the observations made in yeast (FIG. 10) and in insect cells (FIG. 13).

Conjugation of ubiquitin and SUMO to its target proteins is a highly regulated and dynamic process. Several deubiquitinating enzymes (DUBs) have been identified in yeast and other eukaryotic cells (1). Yeast genetics studies show that many of these enzymes are not essential suggesting that an overlapping function is performed by most of these enzymes. DUBs have been most extensively studied and shown to cleave linear ubiquitin fusions as well isopepetide bonds (3, 35). Much less is known about the enzymes that remove SUMO from isopeptide bonds or artificial SUMO-fusion proteins. Hochstrasser and Li have shown that Ulp1 and Ulp2 remove Smt3 and SUMO 1 from proteins and play a role in progression through the G2/M phase and recovery of cells from checkpoint arrest, respectively (20, 21). Ulp1 and Ulp2 cleave C-terminus of SUMO (−GGATY; SEQ ID NO: 59) to mature form (−GG) and de-conjugate Smt3 from the side chains of lysines (20, 21). The sequence similarity of two enzymes is restricted to a 200-amino acid sequence called ULP that contains the catalytically active region. The three-dimensional structure of the ULP domain from Ulp1 has been determined in a complex form with SUMO (Smt3) precursor (27). These studies show that conserved surfaces of SUMO determine the processing and de-conjugation of SUMO. Database searches of the human genome and recent findings suggest that there are at least 7 human ULPs with the size ranging from 238 to 1112 amino acid residues (18, 33, 39). It is intriguing to note that SUMO Ulps are not related to DUBs, suggesting that SUMO Ulps evolved separately from DUBs. The findings that ULP structure is distantly related to adenovirus processing protease, intracellular pathogen Chlammydia trachomatis and other proposed bacterial cystiene protease core domains suggest that this sequence evolved in prokaryotes (20, 21). Detailed properties of the SUMO proteases are provided in described in Table 3.

TABLE 3
SUMO Hydrolases/Proteases
Enzyme	Properties (MW)	Reference
UB1-specific	72 KDa. 6 21 residues	Li and Hochstrasser,
Protease	Cleaves linear fusion and	1999 (REF 20)
ULP1	SUMO isopeptides bonds.
ULP2 (Yeast)	117 KDa, 1034 residues	Li and Hochstrasser,
	Cleaves linear fusions and	2000 (REF 21)
	SUMO isopeptide structures.
SUMO-I C-	30 Kda	Suzuki, et al, 1999
Terminal	Cleaves linear fusions and	(REF 33)
	SUMO isopeptide structures
SUMO-I specific	126 KDa 1112 residues	Kim, et al, 2000
Protease	Specific for SUMO-1 fusion	(REF 18)
SUSP I (Human)	but not Smt3 fusion.
	Does not cleave isopeptide
	bond.
Sentrin specific	All of the SENP enzymes have	Yeh, et al, 2000
Proteases (SENP)	conserved C-terminal region	(REF 39)
SENP1	with core catalytic cysteine.
SENP2	The smallest SENP7 is 238
SENP3	residues and the largest SENP6
SENP4	is 1112 residues.
SENP5
SENP6
SENP7

Ulp1 has proven extremely robust in cleaving a variety of SUMO-fusion proteins expressed in E. coli as described in the present example. We have designed SUMO-GFP fusions in which the N-terminal methionine has been replaced with rest of the 19 amino acids. Attachment of 6×His to N-terminus of SUMO afforded easy purification of the 20 SUMO-GFP fusions from E. coli. The enzyme was active under broad ranges of pH, temperature, salts and imidazole concentration and was very effective in cleaving variety of proteins from SUMO fusion that includes BPTI a 6.49 KDa, Protein G a 7 KDa, β-Glucuronidase (GUS) and 110 KDa β-Galactosidase (GAL) genes. These findings suggest that the Ulp1 is much more robust in cleavage of the SUMO-fusion proteins as compared to its counterpart ubiquitin-fusion hydrolase.

SUMO Protease/Hydrolase is a Robust Enzyme:

Effects of Temperature and Additives

The effects of various additives/conditions and temperature upon the in vitro cleavage reaction were determined as follows: His6smt3MeGFP was expressed from pET24d in Rosetta (DE3) pLysS (Novagen). The recombinant protein was purified as before using Ni-NTA agarose (Qiagen) and then buffer exchanged into 20 mM Tris-HCl pH 8.0, 150 mM NaCl and 5 mM β-mercaptoethanol using a PD-10 column (AP Biotech). Cleavage reactions were performed with 100 ug of the purified protein, 0.5 ul of enzyme, the appropriate amount of a stock solution of additive to generate the final concentrations listed in Table 4, plus the exchange buffer up to a final volume of 200 ul. Reactions were incubated for 1 hour at 37° C. except for those at 4° C. were incubated for 3 hours. The data in FIG. 16 shows that Ulp1 was extremely active at 37° C. as well as at 4° C. Generally, His tagged proteins are purified on nickel columns and eluted with imidazole. We have discovered that the enzyme was remarkably active at 0–300 mM imidazole concentration. The enzyme was highly active at 0.01% SDS and up to 1% triton X 100. See Table 4. Similarly, chaotropic agents such as urea and did not effect the activity of the enzyme up to 2 M. Ulp1 showed 50% activity at 0.5M concentration of guanadinium hydrochloride (FIG. 16 and Table 4). A variety of reagents, including cysteine protease inhibitors, EDTA, PMSF. Pepstatin, Leupeptin, TLCK had no effect on the enzymatic activity (FIG. 17 and Table 4). N-ethymaleimide was active only if incubated with the enzyme prior to addition of the substrate. All the data shown in Table 2 demonstrate that this enzyme is extremely robust and thus constitutes a superior reagent for cleaving fusion proteins under variety of conditions.

TABLE 4

The Effect of Different Conditions on the Ulp1 Hydrolase Activity
Conditions/
Additions        Effect

Environmental:
Temperature      Ulp1 is active over a broad range of temperatures,
                 cleaving from 4 to 37° C.
Salts:
Imidazole        Ulp1 shows similar activity in the range of
                 0 to 300 mM
Detergents:
SDS              0.01% SDS blocks activity
Triton-X         Ulp1 shows similar activity on the range of
                 0 to 0.1%
Chaotrophs
Urea             Ulp1 shows complete activity up to and including a
                 2 M concentration
Gdm HCl          Ulp1 shows 50% activity in 0.5 M but is completely
                 inactive in 1 M concentrations
Protease inhibitors:
E-64             Cysteine protease inhibitor; no affect
EDTA             Metalloprotease inhibitor; no affect
PMSF             Serine protease inhibitor; no affect
Pepstatin        Aspartate protease inhibitor; no affect
Leupeptin        Inhibits serine and cysteine proteases with trypsin-
                 like specificity; no affect
TLCK-HCl         Inhibits serine and cysteine proteases with
                 chymotrypsin-like specificity; no affect
N-ethylmaleimide Cysteine protease inhibitor; on effective if enzyme is
                 preincubated with inhibitor
                 before addition of substrate

Robust Properties of SUMO Hydrolase: Cleavage of Different Size Fusion Proteins Under Broad pH Range:

FIG. 18 shows purification of a 40 kDa MAPKAP2 kinase that was difficult to express unless fused to SUMO. We have shown in Example I (FIG. 8) that this kinase was expressed in a highly soluble form (95%) as fusion to SUMO. FIG. 18 shows that whether purfied from cells expressing at 37° C. or 20° C., the SUMO fusion was efficiently cleaved under the conditions described.

The SUMO hydrolase also functions under broad pH range. FIG. 19 shows kinetics of cleavage at pH 7.5 and 8.0. The data shows that purified SUMO-GFP was completely digested at room temperature. We have also performed experiments from pH 5.5 to 10. The data (not shown) support the notion that this enzyme is active over broad range of pH.

As discussed above, for broad utility of the system it is important that the enzyme be able to cleave fusion proteins of different sizes and structures in vitro. FIG. 20 shows the digestion pattern of SUMO-β-galactosidase (β-Gal) a 110 KDa protein. β-Gal enzyme is composed of tetrameric subunits. The digestion pattern demonstrates that in 20 minutes, SUMO hydrolase was able to cleave 100% of the protein.

Among dozens of proteins expressed as SUMO fusions in our lab, only one, β-GUS, proved partially resistant to cleavage by the hydrolase. Configurations of artificial SUMO fusion are bound to occur wherein the structure of the protein will hinder the ability of the enzyme to recognize and bind the cleavage site of the fusion protein. This problem has been solved by adding small concentrations of urea, which does not inhibit the hydrolase, but results in cleavage the fusion that was previously resistant. FIG. 21 shows the digestion pattern of purified β-GUS and SUMO hydrolase before and after addition of urea. Lane 6 and 9 contain the same amount of SUMO hydrolase to which 2M urea was added during the incubation. Addition of urea allowed complete cleavage of 65 KDa β-GUS in 20 min at room temperature. This data further proves that the SUMO hydrolase cleaves broad spectrum of fusion protein efficiently. Additives such as urea can be added to aid complete cleavage of these structures that are resistant to hydrolase action.

High Throughput Protein Purification of Fusion Proteins: Rapid Peptide Miniprep

We have discovered that, due to the rapid folding properties of SUMO, the fused protein can also be rapidly re-natured after treatment of the crude protein mix with chaotropic agents such as guanidinium hydrochloride or urea. We have developed a simple and rapid procedure to purify SUMO-fused proteins that are expressed in prokaryotes and eukaryotes. This method was tested with SUMO-protein G fusion expressed in E. coli. Cells expressing 6×His-SUMO-G protein fusion were harvested and frozen until required for protein purification. Three times the weight per volume lysis buffer (6 M Guanidinium Chloride, 20 mM Tris-HCl, 150 mM NaCl, pH 8.0) was added to the cell pellet rapidly lyse the cells. The supernatant was loaded onto a pre-equilibrated column containing Ni-NTA agarose (Qiagen), the flow through was collected for analysis. The column was then washed, first with 2 column volumes (CV) of Lysis buffer, followed by 3 CV of wash buffer (20 mM Tris-HCl, 150 mM NaCl 15 mM Imidazole pH 8.0). The fusion protein was then eluted using 2 CV of elution buffer (20 mM Tris-HCl, 150 mM NaCl 300 mM Imidazole pH 8.0). The purified product is present in a native buffer that allows for cleavage and release of the peptide from the Sumo fusion using Ulp1. See FIG. 22. This data demonstrates that it is possible to rapidly purify the fusion protein and cleave it from the resin with Ulp1. It is possible that proteins of higher molecular weights may not rapidly re-nature and be amenable to cleavage by Ulp1. However, since the Ulp1 requires three-dimensional SUMO be intact the purification and cleavage properties are more dependent on the refolding of SUMO. Similar to DNA mini-preps, rapid mini preps for the expression and purification analysis of the fused proteins may be readily employed. Table 5 summarizes the data showing the dramatic enhancement of protein production observed when utilizing the compositions and methods of the present invention. The sequences and vectors utilized in the practice of the invention are shown in FIGS. 23–46.

TABLE 5

Fusion with SUMO Enhances Protein Expression


E. coli Expression       All of the fusion have Met
of UBLs                  N-Termini

SUMO-GFP                 40 fold
Ub-GFP                   40 fold
Urm1-GFP                 50 fold
Hub1-GFP                 2 fold
Rub1-GFP                 50 fold
Apg8-GFP                 40 fold
Apg12-GFP                20 fold
ISG15-GFP                3–5 fold

                         Met and Various N-
Yeast                    Termini

Various UBLs expressed   Copper induction not observed in rich
in rich media.           media, however, Ub, SUMO, ISG15
                         fusions were processed and GFP
                         induced 3–5 fold.
All of the twenty N-     Dramatic induction of GFP following
terminal variants were   fusion with SUMO. At least 50–100
expressed in yeast as    fold induction as compared to unfused
SUMO-X-GFP fusions.      GFP expression. Under current
GFP was processed in all loading conditions (20 ug) GFP was
cases, except when N-    not detectable.
terminal residue was
proline.

Insect Cells             Met as N-termini

SUMO-GFP                 10 fold compared to GFP
gp67-SUMO-GFP            30 fold compared to gp-GFP
gp67-SUMO-GFP            50 fold compared to SUMO-gp67-
                         GFP
Secretion SUMO-GFP       At least 50 fold compared to GFP
Secretion Ub-GFP         At least 50 fold compared to GFP

REFERENCES

• 1. Amerik, A. Y., S. J. Li, and M. Hochstrasser. 2000. Analysis of the deubiquitinating enzymes of the yeast Saccharomyces cerevisiae. Biol Chem 381:981–92.
• 2. Bachmair, A., D. Finley, and A. Varshavsky. 1986. In vivo half-life of a protein is a function of its amino-terminal residue. Science 234:179–86.
• 3. Baker, R. T. 1996. Protein expression using ubiquitin fusion and cleavage. Curr Opin Biotechnol 7:541–6.
• 4. Bayer, P., A. Arndt, S. Metzger, R. Mahajan, F. Melchior, R. Jaenicke, and J. Becker. 1998. Structure determination of the small ubiquitin-related modifier SUMO-1. J Mol Biol 280:275–86.
• 5. Butt, T. R., S. Jonnalagadda, B. P. Monia, E. J. Sternberg, J. A. Marsh, J. M. Stadel, D. J. Ecker, and S. T. Crooke. 1989. Ubiquitin fusion augments the yield of cloned gene products in Escherichia coli. Proc Natl Acad Sci USA 86:2540–4.
• 6. Butt, T. R., E. J. Sternberg, J. A. Gorman, P. Clark, D. Hamer, M. Rosenberg, and S. T. Crooke. 1984. Copper metallothionein of yeast, structure of the gene, and regulation of expression. Proc Natl Acad Sci USA 81:3332–6.
• 7. Ecker, D. J., J. M. Stadel, T. R. Butt, J. A. Marsh, B. P. Monia, D. A. Powers, J. A. Gorman, P. E. Clark, F. Warren, A. Shatzman, and et al. 1989. Increasing gene expression in yeast by fusion to ubiquitin. J Biol Chem 264:7715–9.
• 8. Gietz, D., A. St. Jean, R. A. Woods, and R. H. Schiestl. 1992. Improved method for high efficiency transformation of intact yeast cells. Nucleic Acids Res 20:1425.
• 9. Goward, C. R., J. P. Murphy, T. Atkinson, and D. A. Barstow. 1990. Expression and purification of a truncated recombinant streptococcal protein G. Biochem J 267:171–7.
• 10. Graumann, K., J. L. Wittliff, W. Raffelsberger, L. Miles, A. Jungbauer, and T. R. Butt. 1996. Structural and functional analysis of N-terminal point mutants of the human estrogen receptor. J Steroid Biochem Mol Biol 57:293–300.
• 11. Hicke, L. 1997. Ubiquitin-dependent internalization and down-regulation of plasma membrane proteins. Faseb J 11:1215–26.
• 12. Hochstrasser, M. 2000. Evolution and function of ubiquitin-like protein-conjugation systems. Nat Cell Biol 2:E153–7.
• 13. Hochstrasser, M. 1995. Ubiquitin, proteasomes, and the regulation of intracellular protein degradation. Curr Opin Cell Biol 7:215–23.
• 14. Hochstrasser, M. 1996. Ubiquitin-dependent protein degradation. Annu Rev Genet 30:405–39.
• 15. Jentsch, S., and G. Pyrowolakis. 2000. Ubiquitin and its kin: how close are the family ties? Trends Cell Biol 10:335–42._—00001785_—00001785.
• 16. Johnson, E. S., I. Schwienhorst, R. J. Dohmen, and G. Blobel. 1997. The ubiquitin-like protein Smt3p is activated for conjugation to other proteins by an Aos1p/Uba2p heterodimer. Embo J 16:5509–19.
• 17. Kapust, R. B., and D. S. Waugh. 1999. Escherichia coli maltose-binding protein is uncommonly effective at promoting the solubility of polypeptides to which it is fused. Protein Sci 8:1668–74.
• 18. Kim, K. I., S. H. Baek, Y. J. Jeon, S. Nishimori, T. Suzuki, S. Uchida, N. Shimbara, H. Saitoh, K. Tanaka, and C. H. Chung. 2000. A new SUMO-1-specific protease, SUSP1, that is highly expressed in reproductive organs. J Biol Chem 275:14102–6.
• 19. LaBean, T. H., S. A. Kauffman, and T. R. Butt. 1995. Libraries of random-sequence polypeptides produced with high yield as carboxy-terminal fusions with ubiquitin. Mol Divers 1:29–38.
• 20. Li, S. J., and M. Hochstrasser. 1999. A new protease required for cell-cycle progression in yeast. Nature 398:246–51.
• 21. Li, S. J., and M. Hochstrasser. 2000. The yeast ULP2 (SMT4) gene encodes a novel protease specific for the ubiquitin-like Smt3 protein. Mol Cell Biol 20:2367–77.
• 22. Lyttle, C. R., P. Damian-Matsumura, H. Juul, and T. R. Butt. 1992. Human estrogen receptor regulation in a yeast model system and studies on receptor agonists and antagonists. J Steroid Biochem Mol Biol 42:677–85.
• 23. Mahajan, R., L. Gerace, and F. Melchior. 1998. Molecular characterization of the SUMO-1 modification of RanGAP1 and its role in nuclear envelope association. J Cell Biol 140:259–70.
• 24. Malakhova, O., M. Malakhov, C. Hetherington, and D. E. Zhang. 2002. Lipopolysaccharide activates the expression of ISG15-specific protease UBP43 via interferon regulatory factor 3. J Biol Chem 277:14703–11.
• 25. Marathe, S. V., and J. E. McEwen. 1995. Vectors with the gus reporter gene for identifying and quantitating promoter regions in Saccharomyces cerevisiae. Gene 154:105–7.
• 26. Matunis, M. J., J. Wu, and G. Blobel. 1998. SUMO-1 modification and its role in targeting the Ran GTPase-activating protein, RanGAP1, to the nuclear pore complex. J Cell Biol 140:499–509.
• 27. Mossessova, E., and C. D. Lima. 2000. Ulp1-SUMO crystal structure and genetic analysis reveal conserved interactions and a regulatory element essential for cell growth in yeast. Mol Cell 5:865–76.
• 28. Muller, S., C. Hoege, G. Pyrowolakis, and S. Jentsch. 2001. SUMO, ubiquitin's mysterious cousin. Nat Rev Mol Cell Biol 2:202–10.
• 29. Muller, S., M. J. Matunis, and A. Dejean. 1998. Conjugation with the ubiquitin-related modifier SUMO-1 regulates the partitioning of PML within the nucleus. Embo J 17:61–70.
• 30. Ohsumi, Y. 2001. Molecular dissection of autophagy: two ubiquitin-like systems. Nat Rev Mol Cell Biol 2:211–6.
• 31. Sherman, F., G. Fink, and J. Hicks. 1986. Methods in yeast genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y.
• 32. Sikorski, R. S., and P. Hieter. 1989. A system of shuttle vectors and yeast host strains designed for efficient manipulation of DNA in Saccharomyces cerevisiae. Genetics 122:19–27.
• 33. Suzuki, T., A. Ichiyama, H. Saitoh, T. Kawakami, M. Omata, C. H. Chung, M. Kimura, N. Shimbara, and K. Tanaka. 1999. A new 30-kDa ubiquitin-related SUMO-1 hydrolase from bovine brain. J Biol Chem 274:31131–4.
• 34. Varshavsky, A. 1996. The N-end rule: functions, mysteries, uses. Proc Natl Acad Sci USA 93:12142–9.
• 35. Varshavsky, A. 2000. Ubiquitin fusion technique and its descendants. Methods Enzymol 327:578–93.
• 36. Waldo, G. S., B. M. Standish, J. Bereudzen, and T. C. Terwilliger. 1999. Rapid protein-folding assay using green fluorescent protein. Nat Biotechnol 17:691–5.
• 37. Walfish, P. G., T. Yoganathan, Y. F. Yang, H. Hong, T. R. Butt, and M. R. Stallcup. 1997. Yeast hormone response element assays detect and characterize GRIP1 coactivator-dependent activation of transcription by thyroid and retinoid nuclear receptors. Proc Natl Acad Sci USA 94:3697–702.
• 38. Wright, L. C., J. Seybold, A. Robichaud, I. M. Adcock, and P. J. Barnes. 1998. Phosphodiesterase expression in human epithelial cells. Am J Physiol 275:L694–700.
• 39. Yeh, E. T., L. Gong, and T. Kamitani. 2000. Ubiquitin-like proteins: new wines in new bottles. Gene 248:1–14.

While certain of the preferred embodiments of the present invention have been described and specifically exemplified above, it is not intended that the invention be limited to such embodiments. Various modifications may be made thereto without departing from the scope and spirit of the present invention, as set forth in the following claims.

Claims

1. A kit comprising a recombinant vector containing a nucleic acid sequence encoding small ubiquitin related modifier (sumo) operably linked to a promoter and a multiple cloning site; wherein said multiple cloning site allows for inserting a nucleic acid encoding a protein of interest in-frame and immediately 3′ to the nucleic acid sequence encoding the Gly-Gly cleavage site of sumo.

2. The kit of claim 1, wherein said kit further comprises host cells.

3. The kit of claim 2, wherein said host cells are selected from the group of yeast cells, E. coli, insect cells, and mammalian cells.

4. The kit of claim 1, wherein said kit further comprises reagents for oligonucleotide-based site-directed mutagenesis for altering the nucleic acid encoding said protein of interest such that the altered nucleic acid encodes said protein of interest with an altered amino terminus.

5. The kit of claim 4, wherein said reagents comprise primers for performing oligonucleotide-based site-directed mutagenesis.

6. A kit for purification of a protein from a host cell comprising:

i) a recombinant vector comprising:

a) a nucleic acid sequence encoding sumo;

b) a promoter;

c) a nucleic acid sequence encoding for a purification tag; and

d) a multiple cloning site;

7. The kit of claim 6, wherein said kit further comprises host cells.

8. The kit of claim 7 wherein said host cells are selected from the group of yeast cells, E. coli, insect cells, and mammalian cells.

9. The kit of claim 6 further comprising:

i) a solid support for binding the purification tag,

ii) lysis buffers,

iii) wash buffers,

iv) elution buffers,

v) cleavage buffers, and

vi) instruction material.

10. The kit of claim 1, wherein said sumo is SEQ ID NO: 65.

11. The kit of claim 6, wherein said sumo is SEQ ID NO: 65.

12. The kit of claim 1, wherein said multiple cloning site comprises a Bsa I site.

13. The kit of claim 6, wherein said multiple cloning site comprises a Bsa I site.

14. The kit of claim 1, wherein said multiple.cloning site is nucleotides 426–478 of SEQ ID NO: 37.

15. The kit of claim 6, wherein said multiple cloning site is nucleotides 426–478 of SEQ ID NO: 37.